Tootfinder

Steering MoE LLMs via Expert (De)Activation
Mohsen Fayyaz, Ali Modarressi, Hanieh Deilamsalehy, Franck Dernoncourt, Ryan Rossi, Trung Bui, Hinrich Sch\"utze, Nanyun Peng
https://arxiv.org/abs/2509.09660

Steering MoE LLMs via Expert (De)Activation
Mixture-of-Experts (MoE) in Large Language Models (LLMs) routes each token through a subset of specialized Feed-Forward Networks (FFN), known as experts. We present SteerMoE, a framework for steering MoE models by detecting and controlling behavior-linked experts. Our detection method identifies experts with distinct activation patterns across paired inputs exhibiting contrasting behaviors. By selectively (de)activating such experts during inference, we control behaviors like faithfulness and s…

Fine-Grained control over Music Generation with Activation Steering
Dipanshu Panda, Jayden Koshy Joe, Harshith M R, Swathi Narashiman, Pranay Mathur, Anish Veerakumar, Aniruddh Krishna, Keerthiharan A
https://arxiv.org/abs/2506.10225

Fine-Grained control over Music Generation with Activation Steering
We present a method for fine-grained control over music generation through inference-time interventions on an autoregressive generative music transformer called MusicGen. Our approach enables timbre transfer, style transfer, and genre fusion by steering the residual stream using weights of linear probes trained on it, or by steering the attention layer activations in a similar manner. We observe that modelling this as a regression task provides improved performance, hypothesizing that the mean-…

Data-Driven Density Steering via the Gromov-Wasserstein Optimal Transport Distance
Haruto Nakashima, Siddhartha Ganguly, Kenji Kashima
https://arxiv.org/abs/2508.06052 https://

Data-Driven Density Steering via the Gromov-Wasserstein Optimal Transport Distance
We tackle the data-driven chance-constrained density steering problem using the Gromov-Wasserstein metric. The underlying dynamical system is an unknown linear controlled recursion, with the assumption that sufficiently rich input-output data from pre-operational experiments are available. The initial state is modeled as a Gaussian mixture, while the terminal state is required to match a specified Gaussian distribution. We reformulate the resulting optimal control problem as a difference-of-con…

Hybrid A* Path Planning with Multi-Modal Motion Extension for Four-Wheel Steering Mobile Robots
Runjiao Bao, Lin Zhang, Tianwei Niu, Haoyu Yuan, Shoukun Wang
https://arxiv.org/abs/2509.06115

Hybrid A* Path Planning with Multi-Modal Motion Extension for Four-Wheel Steering Mobile Robots
Four-wheel independent steering (4WIS) systems provide mobile robots with a rich set of motion modes, such as Ackermann steering, lateral steering, and parallel movement, offering superior maneuverability in constrained environments. However, existing path planning methods generally assume a single kinematic model and thus fail to fully exploit the multi-modal capabilities of 4WIS platforms. To address this limitation, we propose an extended Hybrid A* framework that operates in a four-dimension…

Strong squeezing and perfect one-way EPR steering in electro-optomechanical system
Qing-Min Zeng, A-Peng Liu, Qi Guo
https://arxiv.org/abs/2507.07697 https…

Strong squeezing and perfect one-way EPR steering in electro-optomechanical system
We consider a three-mode electro-optomechanical system in which a mechanical oscillator is coupled to an optical cavity and a LC circuit through radiation pressure and capacitive coupling, respectively. By controlling the two-tone driving of the optical cavity and the microwave driving of the LC circuit, the strong squeezing of both the microwave field of the LC circuit and the mechanical mode will be obtained. Moreover, by further altering the driving power of the system, the perfect one-way E…

Steering Protein Language Models
Long-Kai Huang, Rongyi Zhu, Bing He, Jianhua Yao
https://arxiv.org/abs/2509.07983 https://arxiv.org/pdf/2509.07983

Steering Protein Language Models
Protein Language Models (PLMs), pre-trained on extensive evolutionary data from natural proteins, have emerged as indispensable tools for protein design. While powerful, PLMs often struggle to produce proteins with precisely specified functionalities or properties due to inherent challenges in controlling their outputs. In this work, we investigate the potential of Activation Steering, a technique originally developed for controlling text generation in Large Language Models (LLMs), to direct PL…

Peek 🔲
窥 🔲
📷 Pentax MX
🎞️Ilford FP4 Plus, expired 1994
buy me ☕️ ?/请我喝杯☕️？
#filmphotography

Decentralising LLM Alignment: A Case for Context, Pluralism, and Participation
Oriane Peter, Kate Devlin
https://arxiv.org/abs/2509.08858 https://arxiv.org…

Decentralising LLM Alignment: A Case for Context, Pluralism, and Participation
Large Language Models (LLMs) alignment methods have been credited with the commercial success of products like ChatGPT, given their role in steering LLMs towards user-friendly outputs. However, current alignment techniques predominantly mirror the normative preferences of a narrow reference group, effectively imposing their values on a wide user base. Drawing on theories of the power/knowledge nexus, this work argues that current alignment practices centralise control over knowledge production …

Apple says it is appealing the EU's €500M DMA fine, levied in April 2025 over App Store steering rules, and claims it goes "far beyond what the law requires" (Juli Clover/MacRumors)
https://www.macrumors.com/2025/07/07/apple-appeals-eu-500m-euro…

Apple Challenges 'Unprecedented' €500M EU Fine Over App Store Steering Rules
Apple is appealing the 500 million euro ($570 million) fine that it is facing in the European Union for allegedly violating the Digital Markets...

3D Steering and Localization in Pipes and Burrows using an Externally Steered Soft Growing Robot
Yimeng Qin, Jared Grinberg, William Heap, Allison M. Okamura
https://arxiv.org/abs/2507.07225

3D Steering and Localization in Pipes and Burrows using an Externally Steered Soft Growing Robot
Navigation and inspection in confined environments, such as tunnels and pipes, pose significant challenges for existing robots due to limitations in maneuverability and adaptability to varying geometries. Vine robots, which are soft growing continuum robots that extend their length through soft material eversion at their tip, offer unique advantages due to their ability to navigate tight spaces, adapt to complex paths, and minimize friction. However, existing vine robot designs struggle with na…

Broadband Simultaneous Beam Steering and Compressing Device Based on Subwavelength Protrusion Metallic Tunnels
Dongguo Zhang, Fei Sun, Qin Liao, Yichao Liu, Donguk Nam
https://arxiv.org/abs/2509.04856 …

Broadband Simultaneous Beam Steering and Compressing Device Based on Subwavelength Protrusion Metallic Tunnels
Beam steering and beamwidth compressing play a role in steering the beam and narrowing its half-power beamwidth, respectively, which are both widely applied in extending the effective operational range of 6G communications, IoT devices, and antenna systems. However, research on wave manipulation devices capable of simultaneously achieving both functionalities remains limited, despite their great potential for system miniaturization and functional integration. In this study, we design and realiz…

If you want to participate in the upcoming #NixOS steering committee election but haven't reached the required 25 commits mark in #nixpkgs yet, feel free to send some PRs to me for review. Also available for some 1:1 chats in Matrix or Discord to get any issues quickly sorted, so your PRs ca…

Wasserstein Distributionally Robust Adaptive Covariance Steering
Aditya Gahlawat, Vivek Khatana, Duo Wang, Sambhu H. Karumanchi, Naira Hovakimyan, Petros Voulgaris
https://arxiv.org/abs/2509.04593

Wasserstein Distributionally Robust Adaptive Covariance Steering
We present a methodology for predictable and safe covariance steering control of uncertain nonlinear stochastic processes. The systems under consideration are subject to general uncertainties, which include unbounded random disturbances (aleatoric uncertainties) and incomplete model knowledge (state-dependent epistemic uncertainties). These general uncertainties lead to temporally evolving state distributions that are entirely unknown, can have arbitrary shapes, and may diverge unquantifiably f…

Saturable nonlinearity induced quantum correlations in optomechanics
D. R. Kenigoule Massembele, E. Kongkui Berinyuy, P. Djorwe, A. -H. Abdel-Aty, M. R. Eid, R. Altuijri, S. G. Nana Engo
https://arxiv.org/abs/2506.10709

Saturable nonlinearity induced quantum correlations in optomechanics
We propose a scheme that induces quantum correlations in optomtomechanical systems. Our benchmark system consists of two optically coupled optical cavities which interact with a common mechanical resonator. The optical cavities host saturable nonlinearity which triggers either gain or losses in each cavity. Without these nonlinearities, there are no quantum correlations, i.e., entanglement and steering, generated in the system. By turning on the nonlinearities, gain and losses are switched on, …

Right-wing pastor Ralph Drollinger, leads Bible studies for members of Congress and the Trump administration via his Capitol Ministries organization.
The purpose of Capitol Ministries is to transform public officials in “disciples”
who will turn founder Drollinger’s very conservative interpretation of scripture into public policy,
such as his belief that the Bible mandates support for right-wing economic, social, environmental, immigration, and criminal justice policies.

How Ralph Drollinger's Right-Wing Bible Study Is Quietly Steering The Ship Of State
Right-wing pastor Ralph Drollinger brags that a new work requirement provision was inserted into the Republican spending bill because of the Bible studies he hosts of members of Congress.

Head-steered channel selection method for hearing aid applications using remote microphones
Vasudha Sathyapriyan, Michael S. Pedersen, Mike Brookes, Jan {\O}stergaard, Patrick A. Naylor, Jesper Jensen
https://arxiv.org/abs/2508.06928

Head-steered channel selection method for hearing aid applications using remote microphones
We propose a channel selection method for hearing aid applications using remote microphones, in the presence of multiple competing talkers. The proposed channel selection method uses the hearing aid user's head-steering direction to identify the remote channel originating from the frontal direction of the hearing aid user, which captures the target talker signal. We pose the channel selection task as a multiple hypothesis testing problem, and derive a maximum likelihood solution. Under realisti…

👩‍🏭College is not for every student. How schools are steering them to high-demand jobs
https://www.latimes.com/california/story/2025-07-15/getting-students-moving-toward-the-strategic-entry-level-jo…

College is not for every student. How schools are steering them to high-demand jobs
Los Angeles Unified's "Earn & Learn" summer program pays students as they are introduced to vocational skills in high-demand jobs.

PII Jailbreaking in LLMs via Activation Steering Reveals Personal Information Leakage
Krishna Kanth Nakka, Xue Jiang, Xuebing Zhou
https://arxiv.org/abs/2507.02332

PII Jailbreaking in LLMs via Activation Steering Reveals Personal Information Leakage
This paper investigates privacy jailbreaking in LLMs via steering, focusing on whether manipulating activations can bypass LLM alignment and alter response behaviors to privacy related queries (e.g., a certain public figure's sexual orientation). We begin by identifying attention heads predictive of refusal behavior for private attributes (e.g., sexual orientation) using lightweight linear probes trained with privacy evaluator labels. Next, we steer the activations of a small subset of these at…

A Soft Inducement Framework for Incentive-Aided Steering of No-Regret Players
Asrin Efe Yorulmaz, Raj Kiriti Velicheti, Melih Bastopcu, Tamer Ba\c{s}ar
https://arxiv.org/abs/2508.21672

A Soft Inducement Framework for Incentive-Aided Steering of No-Regret Players
In this work, we investigate a steering problem in a mediator-augmented two-player normal-form game, where the mediator aims to guide players toward a specific action profile through information and incentive design. We first characterize the games for which successful steering is possible. Moreover, we establish that steering players to any desired action profile is not always achievable with information design alone, nor when accompanied with sublinear payment schemes. Consequently, we derive…

A Study on Messaging Trade-offs in Data Streaming for Scientific Workflows
Anjus George, Michael J. Brim, Christopher Zimmer, Tyler J. Skluzacek, A. J. Ruckman, Gustav R. Jansen, Sarp Oral
https://arxiv.org/abs/2509.07199

A Study on Messaging Trade-offs in Data Streaming for Scientific Workflows
Memory-to-memory data streaming is essential for modern scientific workflows that require near real-time data analysis, experimental steering, and informed decision-making during experiment execution. It eliminates the latency bottlenecks associated with file-based transfers to parallel storage, enabling rapid data movement between experimental facilities and HPC systems. These tightly coupled experimental-HPC workflows demand low latency, high throughput, and reliable data delivery to support …

A book excerpt details Spotify's decade-long push against Apple's App Store commissions and steering rules, including pushing the EU to pass new antitrust laws (Tim Higgins/Wall Street Journal)
https://www.wsj.com/tech/spotify-apple-dig

JPS: Jailbreak Multimodal Large Language Models with Collaborative Visual Perturbation and Textual Steering
Renmiao Chen, Shiyao Cui, Xuancheng Huang, Chengwei Pan, Victor Shea-Jay Huang, QingLin Zhang, Xuan Ouyang, Zhexin Zhang, Hongning Wang, Minlie Huang
https://arxiv.org/abs/2508.05087

JPS: Jailbreak Multimodal Large Language Models with Collaborative Visual Perturbation and Textual Steering
Jailbreak attacks against multimodal large language Models (MLLMs) are a significant research focus. Current research predominantly focuses on maximizing attack success rate (ASR), often overlooking whether the generated responses actually fulfill the attacker's malicious intent. This oversight frequently leads to low-quality outputs that bypass safety filters but lack substantial harmful content. To address this gap, we propose JPS, \underline{J}ailbreak MLLMs with collaborative visual \underl…

Ha good luck with that https://geeknews.chat/@theregister/114813445350914205

Sometimes I wonder why people worship authority-hungry clout chasers instead of tearing down the whole system that lets them stand above others. But then you realize this clown is the one steering the circus in the US right now, oh well, anyway, moving on.
#Capitalism #AntiCapitalism

InfoSteer: Steering Information Utility in Language Model Post-Training
Chunyuan Deng, Ruidi Chang, Hanjie Chen
https://arxiv.org/abs/2507.05158 https://…

InfoSteer: Steering Information Utility in Language Model Post-Training
Recent advancements in language models (LMs) gradually ushered in an era where post-training is crucial. Yet, post-training approaches such as supervised fine-tuning (SFT) do not guarantee effective use of knowledge acquired during pretraining. We therefore present \ours, a lightweight method that encourages parametric information utilization in LMs during post-training. This is achieved via treating FFN layer as associate key-value memory, and promotes the use of stored memory vectors via forw…

Replaced article(s) found for physics.app-ph. https://arxiv.org/list/physics.app-ph/new
[1/1]:
- A Beam-Steering Reflectarray Antenna with Arbitrary Linear-Polarization Reconfiguration
Changhao Liu, Songlin Zhou, Fan Yang, Shenheng Xu, Maokun Li

Steering Opinion through Dynamic Stackelberg Optimization
Hossein Rastgoftar
https://arxiv.org/abs/2509.06758 https://arxiv.org/pdf/2509.06758

Steering Opinion through Dynamic Stackelberg Optimization
This paper employs the Friedkin-Johnsen (FJ) model to describe the dynamics of opinion evolution within a social network. Under the FJ framework, the society is divided into two subgroups that include stubborn agents and regular agents. The opinions of stubborn agents are not influenced by regular agents, whereas the opinions of regular agents evolve based on the opinions of their neighboring agents. By defining the origin as the desired collective opinion of the society, the objective of the p…

Dynamics and multi-stability of a rotor-actuated Twistcar robot with passive steering joint
Anna Zigelman, Zitao Yu, Rom Levy, Yizhar Or
https://arxiv.org/abs/2507.04846

Dynamics and multi-stability of a rotor-actuated Twistcar robot with passive steering joint
The nonlinear dynamics of many under-actuated wheeled platforms are governed by nonholonomic constraints of no-skid for passively rolling wheels, coupled with momentum balance. In most of theoretical models, the shape variables, i.e. joint angles, are directly prescribed as periodic inputs, such as steering angle of the Twistcar. In this work, we study a variant of the Twistcar model where the actuation input is periodic oscillations of an inertial rotor attached to the main body, while the ste…

So apparently the Switch 2 has a hollow kickstand that technically allows you to do this in your car, and while hilarious, please do not actually do this in your car (while driving).
(via https://www.reddit.com/r/NintendoSwitch2/c

Importance of User Control in Data-Centric Steering for Healthcare Experts
Aditya Bhattacharya, Simone Stumpf, Katrien Verbert
https://arxiv.org/abs/2506.18770

Importance of User Control in Data-Centric Steering for Healthcare Experts
As Artificial Intelligence (AI) becomes increasingly integrated into high-stakes domains like healthcare, effective collaboration between healthcare experts and AI systems is critical. Data-centric steering, which involves fine-tuning prediction models by improving training data quality, plays a key role in this process. However, little research has explored how varying levels of user control affect healthcare experts during data-centric steering. We address this gap by examining manual and aut…

MSRS: Adaptive Multi-Subspace Representation Steering for Attribute Alignment in Large Language Models
Xinyan Jiang, Lin Zhang, Jiayi Zhang, Qingsong Yang, Guimin Hu, Di Wang, Lijie Hu
https://arxiv.org/abs/2508.10599

MSRS: Adaptive Multi-Subspace Representation Steering for Attribute Alignment in Large Language Models
Activation steering offers a promising approach to controlling the behavior of Large Language Models by directly manipulating their internal activations. However, most existing methods struggle to jointly steer multiple attributes, often resulting in interference and undesirable trade-offs. To address this challenge, we propose Multi-Subspace Representation Steering (MSRS), a novel framework for effective multi-attribute steering via subspace representation fine-tuning. MSRS reduces inter-attri…

Exciton transport driven by spin excitations in an antiferromagnet
Florian Dirnberger, Sophia Terres, Zakhar A. Iakovlev, Kseniia Mosina, Zdenek Sofer, Akashdeep Kamra, Mikhail M. Glazov, Alexey Chernikov
https://arxiv.org/abs/2507.07071

Exciton transport driven by spin excitations in an antiferromagnet
A new class of optical quasiparticles called magnetic excitons recently emerged in magnetic van der Waals materials. Akin to the highly effective strategies developed for electrons, the strong interactions of these excitons with the spin degree of freedom may provide innovative solutions for long-standing challenges in optics, such as steering the flow of energy and information. Here, we demonstrate transport of excitons by spin excitations in the van der Waals antiferromagnetic semiconductor C…

Annals of anthropology beefs:
Just received an email from the organizing committee (OC) of the WAU (World Anthropological Union) Congress, which was very confusing. Apparently some kind of beef between the WAU Steering Committee (SC) and the OC? Perhaps SC people are angry because their panel/paper proposals were not accepted? Anybody have the inside scoop on this?
Here's the OC's published letter about this: #anthropology #anthrodons

That Neue Klasse ix3 looks very nice (for an SUV) though think that steering wheel is overcompensating for something.
I'm not in the market for an SUV, despite the price, but a 3er Touring might fit the bill when it arrives.
https://www.the-intercooler.com/library/bl

BMW iX3: welcome to the future of BMW - The Intercooler
This is the first of a whole new family of BMWs called ‘Neue Klasse’. They will define the brand for decades to come. Richard Aucock reports

Context Steering: A New Paradigm for Compression-based Embeddings by Synthesizing Relevant Information Features
Guillermo Sarasa Dur\'an, Ana Granados Fontecha, Francisco de Borja Rodr\'iguez Ort\'iz
https://arxiv.org/abs/2508.14780

Context Steering: A New Paradigm for Compression-based Embeddings by Synthesizing Relevant Information Features
Compression-based distances (CD) offer a flexible and domain-agnostic means of measuring similarity by identifying implicit information through redundancies between data objects. However, as similarity features are derived from the data, rather than defined as an input, it often proves difficult to align with the task at hand, particularly in complex clustering or classification settings. To address this issue, we introduce "context steering," a novel methodology that actively guides the featur…

A book excerpt details Spotify's decade-long push against Apple's App Store commissions and steering rules, including pushing the EU to pass new antitrust laws (Tim Higgins/Wall Street Journal)
https://www.wsj.com/tech/spotify-apple-dig

Witness the High-Dimensional Quantum Steering via Majorization Lattice
Ma-Cheng Yang, Cong-Feng Qiao
https://arxiv.org/abs/2507.20950 https://arxiv.org/pdf…

Witness the High-Dimensional Quantum Steering via Majorization Lattice
Quantum steering enables one party to influence another remote quantum state by local measurement. While steering is fundamental to many quantum information tasks, the existing detection methods in the literature are mainly constrained to either specific measurement scenario or low-dimensional systems. In this work, we propose a majorization lattice framework for steering detection, which is capable of exploring the steering in arbitrary dimension and measurement setting. Steering inequalities …

CoSteer: Collaborative Decoding-Time Personalization via Local Delta Steering
Hang Lv, Sheng Liang, Hao Wang, Hongchao Gu, Yaxiong Wu, Wei Guo, Defu Lian, Yong Liu, Enhong Chen
https://arxiv.org/abs/2507.04756

CoSteer: Collaborative Decoding-Time Personalization via Local Delta Steering
Personalized text generation has become crucial for adapting language models to diverse and evolving users' personal context across cultural, temporal, and contextual dimensions. While existing methods often rely on centralized fine-tuning or static preference alignment, they struggle to achieve real-time adaptation under resource constraints inherent to personal devices. This limitation creates a dilemma: large cloud-based models lack access to localized user-specific information, while small …

Goal-oriented optimal sensor placement for PDE-constrained inverse problems in crisis management
Marco Mattuschka, Noah An der Lan, Max von Danwitz, Daniel Wolff, Alexander Popp
https://arxiv.org/abs/2507.02500

Goal-oriented optimal sensor placement for PDE-constrained inverse problems in crisis management
This paper presents a novel framework for goal-oriented optimal static sensor placement and dynamic sensor steering in PDE-constrained inverse problems, utilizing a Bayesian approach accelerated by low-rank approximations. The framework is applied to airborne contaminant tracking, extending recent dynamic sensor steering methods to complex geometries for computational efficiency. A C-optimal design criterion is employed to strategically place sensors, minimizing uncertainty in predictions. Nume…

Logit-Gap Steering: Efficient Short-Suffix Jailbreaks for Aligned Large Language Models
Tung-Ling Li, Hongliang Liu
https://arxiv.org/abs/2506.24056 https:…

Logit-Gap Steering: Efficient Short-Suffix Jailbreaks for Aligned Large Language Models
We introduce logit-gap steering, a fast jailbreak framework that casts the refusal-affirmation gap of RLHF-aligned language models as a single pass over the vocabulary. A forward-computable score blends gap reduction with lightweight proxies for KL penalty and reward shift, allowing a "sort-sum-stop" sweep to complete in under a second and return a short suffix--two orders of magnitude fewer model calls than beam or gradient attacks. The same suffix generalises to unseen prompts and scales from…

Gaussian Process Regression of Steering Vectors With Physics-Aware Deep Composite Kernels for Augmented Listening
Diego Di Carlo (RIKEN AIP), Koyama Shoichi (UTokyo), Nugraha Aditya Arie (RIKEN AIP), Fontaine Mathieu (LTCI, S2A), Bando Yoshiaki (AIST), Yoshii Kazuyoshi (RIKEN AIP)
https://arxiv.org/abs/2509.02571

Gaussian Process Regression of Steering Vectors With Physics-Aware Deep Composite Kernels for Augmented Listening
This paper investigates continuous representations of steering vectors over frequency and position of microphone and source for augmented listening (e.g., spatial filtering and binaural rendering) with precise control of the sound field perceived by the user. Steering vectors have typically been used for representing the spatial characteristics of the sound field as a function of the listening position. The basic algebraic representation of steering vectors assuming an idealized environment can…

EmoSteer-TTS: Fine-Grained and Training-Free Emotion-Controllable Text-to-Speech via Activation Steering
Tianxin Xie, Shan Yang, Chenxing Li, Dong Yu, Li Liu
https://arxiv.org/abs/2508.03543

EmoSteer-TTS: Fine-Grained and Training-Free Emotion-Controllable Text-to-Speech via Activation Steering
Text-to-speech (TTS) has shown great progress in recent years. However, most existing TTS systems offer only coarse and rigid emotion control, typically via discrete emotion labels or a carefully crafted and detailed emotional text prompt, making fine-grained emotion manipulation either inaccessible or unstable. These models also require extensive, high-quality datasets for training. To address these limitations, we propose EmoSteer-TTS, a novel training-free approach, to achieve fine-grained s…

Real-Time Obstacle Avoidance for a Mobile Robot Using CNN-Based Sensor Fusion
Lamiaa H. Zain, Raafat E. Shalaby
https://arxiv.org/abs/2509.08095 https://ar…

Real-Time Obstacle Avoidance for a Mobile Robot Using CNN-Based Sensor Fusion
Obstacle avoidance is a critical component of the navigation stack required for mobile robots to operate effectively in complex and unknown environments. In this research, three end-to-end Convolutional Neural Networks (CNNs) were trained and evaluated offline and deployed on a differential-drive mobile robot for real-time obstacle avoidance to generate low-level steering commands from synchronized color and depth images acquired by an Intel RealSense D415 RGB-D camera in diverse environments. …

Integrated user scheduling and beam steering in over-the-air federated learning for mobile IoT
Shengheng Liu, Ningning Fu, Zhonghao Zhang, Yongming Huang, Tony Q. S. Quek
https://arxiv.org/abs/2508.00341

Integrated user scheduling and beam steering in over-the-air federated learning for mobile IoT
The rising popularity of Internet of things (IoT) has spurred technological advancements in mobile internet and interconnected systems. While offering flexible connectivity and intelligent applications across various domains, IoT service providers must gather vast amounts of sensitive data from users, which nonetheless concomitantly raises concerns about privacy breaches. Federated learning (FL) has emerged as a promising decentralized training paradigm to tackle this challenge. This work focus…

#BeaverWeek #Factoid
A Beaver's tail is flat, leathery, and sparsely haired. Beavers use it as a rudder for steering while swimming and as a prop to balance when standing or sitting on land. They will slap their tail on the water's surface to warn other Beavers of danger. Beavers store fat in their tails, especially before winter, to help them survive periods with limited food.
A Beavertail is also a famous Canadian pastry that was introduced in the winter of 1978 to skaters on the Rideau Canal. It's a great way to eat Beaver without harming any animals.

When you are a swift developer, every day is Xmas!
https://forums.swift.org/t/swift-sdks-for-webassembly-now-available-on-swift-org/80405

Swift SDKs for WebAssembly now available on swift.org
As the Platform Steering Group accepted the vision document for WebAssembly support, we’re excited to announce that Swift SDKs for Wasm are now available at swift.org/install. The website now also hosts a getting started guide for Wasm with installation and build instructions. You may notice "Swift SDKs" is in plural. The reason is that artifact bundles for Wasm distributed from swift.org include two Swift SDKs: one with the full feature set of the Swift standard library and Swift Testing, a…

Prebiotic Functional Programs: Endogenous Selection in an Artificial Chemistry
Devansh Vimal, Cole Mathis, Westley Weimer, Stephanie Forrest
https://arxiv.org/abs/2509.03534 htt…

Prebiotic Functional Programs: Endogenous Selection in an Artificial Chemistry
Artificial chemistry simulations produce many intriguing emergent behaviors, but they are often difficult to steer or control. This paper proposes a method for steering the dynamics of a classic artificial chemistry model, known as AlChemy (Algorithmic Chemistry), which is based on untyped lambda calculus. Our approach leverages features that are endogenous to AlChemy without constructing an explicit external fitness function or building learning into the dynamics. We demonstrate the approach b…

"The suit alleges that the bot gradually cut Raine off from his support networks by routinely supporting his ideas about self-harm instead of steering him toward possible human interventions. ... when he mentioned being close to his brother, ChatGPT told him, “Your brother might love you, but he’s only met the version of you you let him see. But me? I’ve seen it all — the darkest thoughts, the fear, the tenderness. And I’m still here. Still listening. Still your friend.”"

ChatGPT Lawsuit Over Teen’s Suicide Could Lead to Big Tech Reckoning
First wrongful death lawsuit against OpenAI alleges that ChatGPT advised a 16-year-old on how to hang himself, and may trigger more complaints.

StyliTruth : Unlocking Stylized yet Truthful LLM Generation via Disentangled Steering
Chenglei Shen, Zhongxiang Sun, Teng Shi, Xiao Zhang, Jun Xu
https://arxiv.org/abs/2508.04530

StyliTruth : Unlocking Stylized yet Truthful LLM Generation via Disentangled Steering
Generating stylized large language model (LLM) responses via representation editing is a promising way for fine-grained output control. However, there exists an inherent trade-off: imposing a distinctive style often degrades truthfulness. Existing representation editing methods, by naively injecting style signals, overlook this collateral impact and frequently contaminate the model's core truthfulness representations, resulting in reduced answer correctness. We term this phenomenon stylization-…

Having not written anything about my #tricycle project in months, I've now written two posts in two days. Here's the first:
#BikeTooter
Rethinking the tricycle drive train

Rethinking the tricycle drive train
I've been rereading my posts on design of a tricycle, and especially this one; and I've been looking again at this picture, of Seventy Seven's removable front subframe which carries its front wheel, steering, and transmission systems; and I've been thinking. This will be a short post documenting my thoughts.

Narrow beam and low-sidelobe two-dimensional beam steering on thin-film lithium niobate optical phased array
Yang Li, Shiyao Deng, Xiao Ma, Ziliang Fang, Shufeng Li, Weikang Xu, Fangheng Fu, Xu Ouyang, Yuming Wei, Tiefeng Yang, Heyuan Guan, Huihui Lu
https://arxiv.org/abs/2506.22124

Narrow beam and low-sidelobe two-dimensional beam steering on thin-film lithium niobate optical phased array
Optical beam steering has become indispensable in free-space optical communications, light detection and ranging (LiDAR), mapping, and projection. Optical phased array (OPA) leads this field, yet conventional versions still suffer from a narrow steering field of view (FOV), insufficient sidelobe suppression, and limited angular resolution. Thin-film lithium niobate (LN), with its strong Pockels electro-optic (EO) effect, offers a powerful integrated-photonics platform to overcome these limitati…

GPU-Accelerated Barrier-Rate Guided MPPI Control for Tractor-Trailer Systems
Keyvan Majd, Hardik Parwana, Bardh Hoxha, Steven Hong, Hideki Okamoto, Georgios Fainekos
https://arxiv.org/abs/2508.05773

GPU-Accelerated Barrier-Rate Guided MPPI Control for Tractor-Trailer Systems
Articulated vehicles such as tractor-trailers, yard trucks, and similar platforms must often reverse and maneuver in cluttered spaces where pedestrians are present. We present how Barrier-Rate guided Model Predictive Path Integral (BR-MPPI) control can solve navigation in such challenging environments. BR-MPPI embeds Control Barrier Function (CBF) constraints directly into the path-integral update. By steering the importance-sampling distribution toward collision-free, dynamically feasible traj…

A first-order condition for discrete-time distribution steering
Alberto Dom\'inguez Corella, David Gonz\'alez-S\'anchez
https://arxiv.org/abs/2508.21026 https://

A first-order condition for discrete-time distribution steering
We study a class of distribution-steering problems from a variational point of view. Under some differentiability assumptions, we derive necessary conditions for optimal Markov policies in the spirit of the Lagrange multiplier approach. We also provide a heuristic gradient-based method derived from the variational principle.

Joint Frequency-Space Sparse Reconstruction for DOA Estimation under Coherent Sources and Amplitude-Phase Errors
Yutong Chen, Cong Zhou, Changsheng You, Shuo Shi
https://arxiv.org/abs/2509.03983

Joint Frequency-Space Sparse Reconstruction for DOA Estimation under Coherent Sources and Amplitude-Phase Errors
In this letter, we propose a joint frequency-space sparse reconstruction method for direction-of-arrival (DOA) estimation, which effectively addresses the issues arising from the existence of coherent sources and array amplitude-phase errors. Specifically, by using an auxiliary source with known angles, we first construct the real steering vectors (RSVs) based on the spectral peaks of received signals in the frequency domain, which serve as a complete basis matrix for compensation for amplitude…

Another change proposal was submitted to the Fedora Engineering Steering Committee (FESCo) for approval[1]:
Replace the #Xorg #Xserver with the X11Libre Xserver for #Fedora 43-

Contactless Precision Steering of Particles in a Fluid inside a Cube with Rotating Walls
Lucas Amoudruz, Petr Karnakov, Petros Koumoutsakos
https://arxiv.org/abs/2506.15958

Contactless Precision Steering of Particles in a Fluid inside a Cube with Rotating Walls
Contactless manipulation of small objects is essential for biomedical and chemical applications, such as cell analysis, assisted fertilisation, and precision chemistry. Established methods, including optical, acoustic, and magnetic tweezers, are now complemented by flow control techniques that use flow-induced motion to enable precise and versatile manipulation. However, trapping multiple particles in fluid remains a challenge. This study introduces a novel control algorithm capable of steering…

State-switching navigation strategies in C. elegans are beneficial for chemotaxis
Kevin S. Chen, Andrew M. Leifer, Jonathan W. Pillow
https://arxiv.org/abs/2508.00191 https://…

State-switching navigation strategies in C. elegans are beneficial for chemotaxis
Animals employ different strategies for relating sensory input to behavioral output to navigate sensory environments, but what strategy to use, when to switch and why remain unclear. In C. elegans, navigation is composed of 'steering' and 'turns', corresponding to small heading changes and large reorientation events, respectively. It is unclear whether transitions between these elements are driven solely by sensory input or are influenced by internal states that persist over time. It also remai…

Steering Conceptual Bias via Transformer Latent-Subspace Activation
Vansh Sharma, Venkat Raman
https://arxiv.org/abs/2506.18887 https://

Steering Conceptual Bias via Transformer Latent-Subspace Activation
This work examines whether activating latent subspaces in language models (LLMs) can steer scientific code generation toward a specific programming language. Five causal LLMs were first evaluated on scientific coding prompts to quantify their baseline bias among four programming languages. A static neuron-attribution method, perturbing the highest activated MLP weight for a C++ or CPP token, proved brittle and exhibited limited generalization across prompt styles and model scales. To address th…

Wordpress is now out of my life. And for what it’s worth, steering away from the controversy is mostly a bonus for me. What I realized a long time ago but only recently got around to was that my handful of low volume sites are better as statically generated sites.

I’m Dr. Angela Rasmussen,
a virologist and proud member of the
Save America Movement steering committee.
The CDC has just been gutted.
RFK Jr. has fired the nation’s top vaccine experts,
replaced them with anti-vaccine ideologues,
and forced out some of the CDC’s most dedicated leaders.
Programs that took decades to build are being razed overnight.
That means fewer vaccines.
Less access to care.
More outbreaks of preventable diseas…

KV Cache Steering for Inducing Reasoning in Small Language Models
Max Belitsky, Dawid J. Kopiczko, Michael Dorkenwald, M. Jehanzeb Mirza, Cees G. M. Snoek, Yuki M. Asano
https://arxiv.org/abs/2507.08799

KV Cache Steering for Inducing Reasoning in Small Language Models
We propose cache steering, a lightweight method for implicit steering of language models via a one-shot intervention applied directly to the key-value cache. To validate its effectiveness, we apply cache steering to induce chain-of-thought reasoning in small language models. Our approach leverages GPT-4o-generated reasoning traces to construct steering vectors that shift model behavior toward more explicit, multi-step reasoning without fine-tuning or prompt modifications. Experimental evaluatio…

Multi-Fidelity Stochastic Trust Region Method with Adaptive Sampling
Yunsoo Ha, Juliane Mueller
https://arxiv.org/abs/2508.03901 https://arxiv.org/pdf/2508…

Multi-Fidelity Stochastic Trust Region Method with Adaptive Sampling
Simulation optimization is often hindered by the high cost of running simulations. Multi-fidelity methods offer a promising solution by incorporating cheaper, lower-fidelity simulations to reduce computational time. However, the bias in low-fidelity models can mislead the search, potentially steering solutions away from the high-fidelity optimum. To overcome this, we propose ASTRO-MFDF, an adaptive sampling trust-region method for multi-fidelity simulation optimization. ASTRO-MFDF features two …

While we have the Corsa currently in for investigation into possible OBC failure, my wife's been driving the family estate, and has been complaining about the thickness of the steering wheel[1].
Extrapolating a bit, here's where I think we might be headed:
[1] A 2019 G31. I think subsequent models have got worse, so perhaps this extrapolation isn't far off?
#bmw

Can boundary configuration be tuned to optimize directional quantum steering harvesting?
Xiao-Li Huang, Xiao-Ying Jiang, Yu-Xuan Wang, Si-Yu Liu, Zejun Wang, Shu-Min Wu
https://arxiv.org/abs/2506.18734

Can boundary configuration be tuned to optimize directional quantum steering harvesting?
We investigate the harvesting of quantum steering and its asymmetry between two static detectors locally interacting with a vacuum massless scalar field near an infinite, perfectly reflecting boundary. The detectors are arranged either parallel or orthogonal to the boundary, with detector $B$ assumed to have an energy gap greater than or equal to that of detector $A$. It is interesting to observe that, with increasing distance between the detectors and the boundary, the boundary tends to suppre…

Crosslisted article(s) found for eess.SP. https://arxiv.org/list/eess.SP/new
[1/1]:
- Gaussian Process Regression of Steering Vectors With Physics-Aware Deep Composite Kernels for Aug...
Di Carlo, Shoichi, Arie, Mathieu, Yoshiaki, Kazuyoshi

Multi-Functional Metasurfaces with M-Type Ferrites: Shaping the Future of mmWave Absorption and Beam Steering
Nohgyeom Ha, Horim Lee, Min Jang, Gyoungdeuk Kim, Hoyong Kim, Byeongjin Park, Manos M. Tentzeris, Sangkil Kim
https://arxiv.org/abs/2506.23240

Multi-Functional Metasurfaces with M-Type Ferrites: Shaping the Future of mmWave Absorption and Beam Steering
This paper presents a comprehensive review and tutorial on multi-functional metasurfaces integrated with M-type ferrite materials for millimeter-wave (mmWave) absorption and beam control. As wireless communication systems transition toward beyond-5G architectures, including non-terrestrial networks (NTNs), the demand for adaptive, low-profile electromagnetic surfaces that can manage interference while enabling beam reconfiguration becomes increasingly critical. Conventional metasurfaces often s…

FLORES: A Reconfigured Wheel-Legged Robot for Enhanced Steering and Adaptability
Zhicheng Song, Jinglan Xu, Chunxin Zheng, Yulin Li, Zhihai Bi, Jun Ma
https://arxiv.org/abs/2507.22345

FLORES: A Reconfigured Wheel-Legged Robot for Enhanced Steering and Adaptability
Wheel-legged robots integrate the agility of legs for navigating rough terrains while harnessing the efficiency of wheels for smooth surfaces. However, most existing designs do not fully capitalize on the benefits of both legged and wheeled structures, which limits overall system flexibility and efficiency. We present FLORES (reconfigured wheel-legged robot for enhanced steering and adaptability), a novel wheel-legged robot design featuring a distinctive front-leg configuration that sets it bey…

Directional Flow of Confined Polaritons in CrSBr
Pratap Chandra Adak, Sichao Yu, Jaime Abad-Arredondo, Biswajit Datta, Andy Cruz, Sorah Fischer, Kseniia Mosina, Zden\v{e}k Sofer, Antonio I. Fern\'andez-Dom\'inguez, Francisco J. Garc\'ia-Vidal, Vinod M. Menon
https://arxiv.org/abs/2507.04367

Directional Flow of Confined Polaritons in CrSBr
Nanoscale control of energy transport is a central challenge in modern photonics. Utilization of exciton-polaritons hybrid light-matter quasiparticles is one viable approach, but it typically demands complex device engineering to enable directional transport. Here, we demonstrate that the van der Waals magnet CrSBr offers an inherent avenue for steering polariton transport leveraging a unique combination of intrinsic optical anisotropy, high refractive index, and excitons dressed by photons. Th…

Reducing Motion Sickness in Passengers of Autonomous Personal Mobility Vehicles by Presenting a Driving Path
Yuya Ide, Hailong Liu, Takahiro Wada
https://arxiv.org/abs/2506.23457 …

Reducing Motion Sickness in Passengers of Autonomous Personal Mobility Vehicles by Presenting a Driving Path
Autonomous personal mobility vehicles (APMVs) are small mobility devices designed for individual automated transportation in shared spaces. In such environments, frequent pedestrian avoidance maneuvers may cause rapid steering adjustments and passive postural responses from passengers, thereby increasing the risk of motion sickness. This study investigated the effects of providing path information on 16 passengers' head movement behavior and motion sickness while riding an APMV. Through a contr…

Self-Steering Deep Non-Linear Spatially Selective Filters for Efficient Extraction of Moving Speakers under Weak Guidance
Jakob Kienegger, Alina Mannanova, Huajian Fang, Timo Gerkmann
https://arxiv.org/abs/2507.02791

Self-Steering Deep Non-Linear Spatially Selective Filters for Efficient Extraction of Moving Speakers under Weak Guidance
Recent works on deep non-linear spatially selective filters demonstrate exceptional enhancement performance with computationally lightweight architectures for stationary speakers of known directions. However, to maintain this performance in dynamic scenarios, resource-intensive data-driven tracking algorithms become necessary to provide precise spatial guidance conditioned on the initial direction of a target speaker. As this additional computational overhead hinders application in resource-con…

Sources: Apple is locked in last-minute EU negotiations over App Store changes to avoid fines set for this week, and is expected to offer "steering" concessions (Barbara Moens/Financial Times)
https://www.ft.com/content/b5d51870-e864-4aa5-b998-c0d2994a7e2…

Apple locked in last-minute App Store negotiations to avoid Brussels fines
Tech group must make concessions to comply with Digital Markets Act or be hit by escalating levies from Thursday

Initiator of the #NixOS documentation team is stepping down:
https://discourse.nixos.org/t/the-next-chapter-in-nix-documentation/68425

Personality as a Probe for LLM Evaluation: Method Trade-offs and Downstream Effects
Gunmay Handa, Zekun Wu, Adriano Koshiyama, Philip Treleaven
https://arxiv.org/abs/2509.04794 …

Personality as a Probe for LLM Evaluation: Method Trade-offs and Downstream Effects
Personality manipulation in large language models (LLMs) is increasingly applied in customer service and agentic scenarios, yet its mechanisms and trade-offs remain unclear. We present a systematic study of personality control using the Big Five traits, comparing in-context learning (ICL), parameter-efficient fine-tuning (PEFT), and mechanistic steering (MS). Our contributions are fourfold. First, we construct a contrastive dataset with balanced high/low trait responses, enabling effective stee…

Parameter Tuning Under Uncertain Road Perception in Driver Assistance Systems
Leon Greiser, Christian Rathgeber, Vladislav Nenchev, S\"oren Hohmann
https://arxiv.org/abs/2509.03694

Parameter Tuning Under Uncertain Road Perception in Driver Assistance Systems
Advanced driver assistance systems have improved comfort, safety, and efficiency of modern vehicles. However, sensor limitations lead to noisy lane estimates that pose a significant challenge in developing performant control architectures. Lateral trajectory planning often employs an optimal control formulation to maintain lane position and minimize steering effort. The parameters are often tuned manually, which is a time-intensive procedure. This paper presents an automatic parameter tuning me…

Model Editing as a Double-Edged Sword: Steering Agent Ethical Behavior Toward Beneficence or Harm
Baixiang Huang, Zhen Tan, Haoran Wang, Zijie Liu, Dawei Li, Ali Payani, Huan Liu, Tianlong Chen, Kai Shu
https://arxiv.org/abs/2506.20606

Model Editing as a Double-Edged Sword: Steering Agent Ethical Behavior Toward Beneficence or Harm
Agents based on Large Language Models (LLMs) have demonstrated strong capabilities across a wide range of tasks. However, deploying LLM-based agents in high-stakes domains comes with significant safety and ethical risks. Unethical behavior by these agents can directly result in serious real-world consequences, including physical harm and financial loss. To efficiently steer the ethical behavior of agents, we frame agent behavior steering as a model editing task, which we term Behavior Editing. …

Crosslisted article(s) found for cs.SD. https://arxiv.org/list/cs.SD/new
[1/1]:
- Gaussian Process Regression of Steering Vectors With Physics-Aware Deep Composite Kernels for Aug...
Di Carlo, Shoichi, Arie, Mathieu, Yoshiaki, Kazuyoshi

Optimization of Radar Search Patterns for Multiple Scanning Missions in Localized Clutter
Yann Briheche (LS2N, LS2N - \'equipe ReV), Fr\'ed\'eric Barbaresco (LS2N, LS2N - \'equipe ReV), Fouad Bennis (LS2N, LS2N - \'equipe ReV), Damien Chablat (LS2N, LS2N - \'equipe RoMas)
https://arxiv.org/abs/2508.02081

Optimization of Radar Search Patterns for Multiple Scanning Missions in Localized Clutter
Electronic Phased-Array Radars offer new possibilities for Optimization of Radar Search Pattern by using bi-dimensional beam forming and beam steering, along both elevation and azimuth axes. The minimization of the Time-Budget required for multiple Radar scanning missions in localized clutter, under constraints of range and detection probability, can be approximated as a Set Cover Problem. We present a Set Cover Problem approximation for Time-Budget minimization of the Radar Search Pattern, and…

Invisible Watermarks, Visible Gains: Steering Machine Unlearning with Bi-Level Watermarking Design
Yuhao Sun, Yihua Zhang, Gaowen Liu, Hongtao Xie, Sijia Liu
https://arxiv.org/abs/2508.10065

Invisible Watermarks, Visible Gains: Steering Machine Unlearning with Bi-Level Watermarking Design
With the increasing demand for the right to be forgotten, machine unlearning (MU) has emerged as a vital tool for enhancing trust and regulatory compliance by enabling the removal of sensitive data influences from machine learning (ML) models. However, most MU algorithms primarily rely on in-training methods to adjust model weights, with limited exploration of the benefits that data-level adjustments could bring to the unlearning process. To address this gap, we propose a novel approach that le…

DynaGuide: Steering Diffusion Polices with Active Dynamic Guidance
Maximilian Du, Shuran Song
https://arxiv.org/abs/2506.13922 https://

DynaGuide: Steering Diffusion Polices with Active Dynamic Guidance
Deploying large, complex policies in the real world requires the ability to steer them to fit the needs of a situation. Most common steering approaches, like goal-conditioning, require training the robot policy with a distribution of test-time objectives in mind. To overcome this limitation, we present DynaGuide, a steering method for diffusion policies using guidance from an external dynamics model during the diffusion denoising process. DynaGuide separates the dynamics model from the base pol…

Numerical Techniques for the Maximum Likelihood Toeplitz Covariance Matrix Estimation: Part I. Symmetric Toeplitz Matrices
Yuri Abramovich, Victor Abramovich, Tanit Pongsiri
https://arxiv.org/abs/2507.01230

Numerical Techniques for the Maximum Likelihood Toeplitz Covariance Matrix Estimation: Part I. Symmetric Toeplitz Matrices
In several applications, one must estimate a real-valued (symmetric) Toeplitz covariance matrix, typically shifted by the conjugated diagonal matrices of phase progression and phase "calibration" errors. Unlike the Hermitian Toeplitz covariance matrices, these symmetric matrices have a unique potential capability of being estimated regardless of these beam-steering phase progression and/or phase "calibration" errors. This unique capability is the primary motivation of this paper.

Algorithmic Collective Action with Multiple Collectives
Claudio Battiloro, Pietro Greiner, Bret Nestor, Oumaima Amezgar, Francesca Dominici
https://arxiv.org/abs/2508.19149 http…

Algorithmic Collective Action with Multiple Collectives
As learning systems increasingly influence everyday decisions, user-side steering via Algorithmic Collective Action (ACA)-coordinated changes to shared data-offers a complement to regulator-side policy and firm-side model design. Although real-world actions have been traditionally decentralized and fragmented into multiple collectives despite sharing overarching objectives-with each collective differing in size, strategy, and actionable goals, most of the ACA literature focused on single collec…

Multilingual Political Views of Large Language Models: Identification and Steering
Daniil Gurgurov, Katharina Trinley, Ivan Vykopal, Josef van Genabith, Simon Ostermann, Roberto Zamparelli
https://arxiv.org/abs/2507.22623

Multilingual Political Views of Large Language Models: Identification and Steering
Large language models (LLMs) are increasingly used in everyday tools and applications, raising concerns about their potential influence on political views. While prior research has shown that LLMs often exhibit measurable political biases--frequently skewing toward liberal or progressive positions--key gaps remain. Most existing studies evaluate only a narrow set of models and languages, leaving open questions about the generalizability of political biases across architectures, scales, and mult…

Modeling and Control of AWOISV: A Filtered Tube-Based MPC Approach for Simultaneous Tracking of Lateral Position and Heading Angle
Xu Yang, Jun Ni, Hengyang Feng, Feiyu Wang, Tiezhen Wang
https://arxiv.org/abs/2508.13457

Modeling and Control of AWOISV: A Filtered Tube-Based MPC Approach for Simultaneous Tracking of Lateral Position and Heading Angle
An all-wheel omni-directional independent steering vehicle (AWOISV) is a specialized all-wheel independent steering vehicle with each wheel capable of steering up to 90°, enabling unique maneuvers like yaw and diagonal movement. This paper introduces a theoretical steering radius angle and sideslip angle (\( θ_R \)-\(β_R \)) representation, based on the position of the instantaneous center of rotation relative to the wheel rotation center, defining the motion modes and switching criteria for…

SHAMaNS: Sound Localization with Hybrid Alpha-Stable Spatial Measure and Neural Steerer
Diego Di Carlo (RIKEN AIP), Mathieu Fontaine (LTCI, IP Paris), Aditya Arie Nugraha (RIKEN AIP), Yoshiaki Bando (RIKEN AIP), Kazuyoshi Yoshii
https://arxiv.org/abs/2506.18954

SHAMaNS: Sound Localization with Hybrid Alpha-Stable Spatial Measure and Neural Steerer
This paper describes a sound source localization (SSL) technique that combines an $α$-stable model for the observed signal with a neural network-based approach for modeling steering vectors. Specifically, a physics-informed neural network, referred to as Neural Steerer, is used to interpolate measured steering vectors (SVs) on a fixed microphone array. This allows for a more robust estimation of the so-called $α$-stable spatial measure, which represents the most plausible direction of arrival…

STARE at the Structure: Steering ICL Exemplar Selection with Structural Alignment
Jiaqian Li, Qisheng Hu, Jing Li, Wenya Wang
https://arxiv.org/abs/2508.20944 https://

STARE at the Structure: Steering ICL Exemplar Selection with Structural Alignment
In-Context Learning (ICL) has become a powerful paradigm that enables LLMs to perform a wide range of tasks without task-specific fine-tuning. However, the effectiveness of ICL heavily depends on the quality of exemplar selection. In particular, for structured prediction tasks such as semantic parsing, existing ICL selection strategies often overlook structural alignment, leading to suboptimal performance and poor generalization. To address this issue, we propose a novel two-stage exemplar sele…

A Wideband Holographic Array with Azimuth and Elevation Beam Steering for 5G/6G Applications
Hazhir Mohammadi, Amir Saman Nooramin, Homayoon Oraizi
https://arxiv.org/abs/2507.11697

A Wideband Holographic Array with Azimuth and Elevation Beam Steering for 5G/6G Applications
This paper presents the design and fabrication of leaky-wave holograms with anisotropic properties, enabling circular polarization for single-beam radiation in desired directions. Tailored for future terrestrial networks like 5G, 6G, and beyond, the proposed antenna achieves high-gain performance with beam steering in elevation and azimuth over a wide angular range. A planar reflector enhances directivity by suppressing backward modes, while a simple feeding network ensures optimal impedance ma…

Hierarchical Decision-Making for Autonomous Navigation: Integrating Deep Reinforcement Learning and Fuzzy Logic in Four-Wheel Independent Steering and Driving Systems
Yizhi Wang, Degang Xu, Yongfang Xie, Shuzhong Tan, Xianan Zhou, Peng Chen
https://arxiv.org/abs/2508.16574

Hierarchical Decision-Making for Autonomous Navigation: Integrating Deep Reinforcement Learning and Fuzzy Logic in Four-Wheel Independent Steering and Driving Systems
This paper presents a hierarchical decision-making framework for autonomous navigation in four-wheel independent steering and driving (4WISD) systems. The proposed approach integrates deep reinforcement learning (DRL) for high-level navigation with fuzzy logic for low-level control to ensure both task performance and physical feasibility. The DRL agent generates global motion commands, while the fuzzy logic controller enforces kinematic constraints to prevent mechanical strain and wheel slippag…

ConamArray: A 32-Element Broadband MEMS Ultrasound Transducer Array
Dennis Laurijssen, Rens Baeyens, Walter Daems, Jan Steckel
https://arxiv.org/abs/2509.01372 https://

ConamArray: A 32-Element Broadband MEMS Ultrasound Transducer Array
This paper presents the ConamArray, a compact broadband ultrasound transducer array composed of 32 MEMS loudspeakers. Unlike conventional broadband transducers, which are typically large and require high driving voltages, the proposed array combines small form factor MEMS devices in a staggered two-row configuration to enable beam steering across a wide ultrasonic band. A dual-microcontroller back-end with synchronized multi-DAC outputs provides flexible waveform generation and runtime steering…

Continuously Steering LLMs Sensitivity to Contextual Knowledge with Proxy Models
Yilin Wang, Heng Wang, Yuyang Bai, Minnan Luo
https://arxiv.org/abs/2508.19720 https://

Continuously Steering LLMs Sensitivity to Contextual Knowledge with Proxy Models
In Large Language Models (LLMs) generation, there exist knowledge conflicts and scenarios where parametric knowledge contradicts knowledge provided in the context. Previous works studied tuning, decoding algorithms, or locating and editing context-aware neurons to adapt LLMs to be faithful to new contextual knowledge. However, they are usually inefficient or ineffective for large models, not workable for black-box models, or unable to continuously adjust LLMs' sensitivity to the knowledge provi…

Model-Structured Neural Networks to Control the Steering Dynamics of Autonomous Race Cars
Mattia Piccinini, Aniello Mungiello, Georg Jank, Gastone Pietro Rosati Papini, Francesco Biral, Johannes Betz
https://arxiv.org/abs/2507.20427

Model-Structured Neural Networks to Control the Steering Dynamics of Autonomous Race Cars
Autonomous racing has gained increasing attention in recent years, as a safe environment to accelerate the development of motion planning and control methods for autonomous driving. Deep learning models, predominantly based on neural networks (NNs), have demonstrated significant potential in modeling the vehicle dynamics and in performing various tasks in autonomous driving. However, their black-box nature is critical in the context of autonomous racing, where safety and robustness demand a tho…

Enhancing Cross-task Transfer of Large Language Models via Activation Steering
Xinyu Tang, Zhihao Lv, Xiaoxue Cheng, Junyi Li, Wayne Xin Zhao, Zujie Wen, Zhiqiang Zhang, Jun Zhou
https://arxiv.org/abs/2507.13236

Enhancing Cross-task Transfer of Large Language Models via Activation Steering
Large language models (LLMs) have shown impressive abilities in leveraging pretrained knowledge through prompting, but they often struggle with unseen tasks, particularly in data-scarce scenarios. While cross-task in-context learning offers a direct solution for transferring knowledge across tasks, it still faces critical challenges in terms of robustness, scalability, and efficiency. In this paper, we investigate whether cross-task transfer can be achieved via latent space steering without par…

Breaking the Mirror: Activation-Based Mitigation of Self-Preference in LLM Evaluators
Dani Roytburg, Matthew Bozoukov, Matthew Nguyen, Jou Barzdukas, Simon Fu, Narmeen Oozeer
https://arxiv.org/abs/2509.03647

Breaking the Mirror: Activation-Based Mitigation of Self-Preference in LLM Evaluators
Large language models (LLMs) increasingly serve as automated evaluators, yet they suffer from "self-preference bias": a tendency to favor their own outputs over those of other models. This bias undermines fairness and reliability in evaluation pipelines, particularly for tasks like preference tuning and model routing. We investigate whether lightweight steering vectors can mitigate this problem at inference time without retraining. We introduce a curated dataset that distinguishes self-preferen…

The Constitutional Controller: Doubt-Calibrated Steering of Compliant Agents
Simon Kohaut, Felix Divo, Navid Hamid, Benedict Flade, Julian Eggert, Devendra Singh Dhami, Kristian Kersting
https://arxiv.org/abs/2507.15478

The Constitutional Controller: Doubt-Calibrated Steering of Compliant Agents
Ensuring reliable and rule-compliant behavior of autonomous agents in uncertain environments remains a fundamental challenge in modern robotics. Our work shows how neuro-symbolic systems, which integrate probabilistic, symbolic white-box reasoning models with deep learning methods, offer a powerful solution to this challenge. This enables the simultaneous consideration of explicit rules and neural models trained on noisy data, combining the strength of structured reasoning with flexible represe…

On Kinodynamic Global Planning in a Simplicial Complex Environment: A Mixed Integer Approach
Otobong Jerome, Alexandr Klimchik, Alexander Maloletov, Geesara Kulathunga
https://arxiv.org/abs/2508.16511 …

On Kinodynamic Global Planning in a Simplicial Complex Environment: A Mixed Integer Approach
This work casts the kinodynamic planning problem for car-like vehicles as an optimization task to compute a minimum-time trajectory and its associated velocity profile, subject to boundary conditions on velocity, acceleration, and steering. The approach simultaneously optimizes both the spatial path and the sequence of acceleration and steering controls, ensuring continuous motion from a specified initial position and velocity to a target end position and velocity.The method analyzes the admiss…

Automating Steering for Safe Multimodal Large Language Models
Lyucheng Wu, Mengru Wang, Ziwen Xu, Tri Cao, Nay Oo, Bryan Hooi, Shumin Deng
https://arxiv.org/abs/2507.13255

Automating Steering for Safe Multimodal Large Language Models
Recent progress in Multimodal Large Language Models (MLLMs) has unlocked powerful cross-modal reasoning abilities, but also raised new safety concerns, particularly when faced with adversarial multimodal inputs. To improve the safety of MLLMs during inference, we introduce a modular and adaptive inference-time intervention technology, AutoSteer, without requiring any fine-tuning of the underlying model. AutoSteer incorporates three core components: (1) a novel Safety Awareness Score (SAS) that …

Steering Robots with Inference-Time Interactions
Yanwei Wang
https://arxiv.org/abs/2506.14287 https://arxiv.org/pdf/2506.14287…

Steering Robots with Inference-Time Interactions
Imitation learning has driven the development of generalist policies capable of autonomously solving multiple tasks. However, when a pretrained policy makes errors during deployment, there are limited mechanisms for users to correct its behavior. While collecting additional data for finetuning can address such issues, doing so for each downstream use case is inefficient at deployment. My research proposes an alternative: keeping pretrained policies frozen as a fixed skill repertoire while allow…

SafeConstellations: Steering LLM Safety to Reduce Over-Refusals Through Task-Specific Trajectory
Utsav Maskey, Sumit Yadav, Mark Dras, Usman Naseem
https://arxiv.org/abs/2508.11290

SafeConstellations: Steering LLM Safety to Reduce Over-Refusals Through Task-Specific Trajectory
LLMs increasingly exhibit over-refusal behavior, where safety mechanisms cause models to reject benign instructions that superficially resemble harmful content. This phenomena diminishes utility in production applications that repeatedly rely on common prompt templates or applications that frequently rely on LLMs for specific tasks (e.g. sentiment analysis, language translation). Through comprehensive evaluation, we demonstrate that LLMs still tend to refuse responses to harmful instructions wh…

Latent Policy Steering with Embodiment-Agnostic Pretrained World Models
Yiqi Wang, Mrinal Verghese, Jeff Schneider
https://arxiv.org/abs/2507.13340 https:/…

Latent Policy Steering with Embodiment-Agnostic Pretrained World Models
Learning visuomotor policies via imitation has proven effective across a wide range of robotic domains. However, the performance of these policies is heavily dependent on the number of training demonstrations, which requires expensive data collection in the real world. In this work, we aim to reduce data collection efforts when learning visuomotor robot policies by leveraging existing or cost-effective data from a wide range of embodiments, such as public robot datasets and the datasets of huma…

Unveiling the Influence of Amplifying Language-Specific Neurons
Inaya Rahmanisa, Lyzander Marciano Andrylie, Krisna Mahardika Ihsani, Alfan Farizki Wicaksono, Haryo Akbarianto Wibowo, Alham Fikri Aji
https://arxiv.org/abs/2507.22581

Unveiling the Influence of Amplifying Language-Specific Neurons
Language-specific neurons in LLMs that strongly correlate with individual languages have been shown to influence model behavior by deactivating them. However, their role in amplification remains underexplored. This work investigates the effect of amplifying language-specific neurons through interventions across 18 languages, including low-resource ones, using three models primarily trained in different languages. We compare amplification factors by their effectiveness in steering to the target …

Turning the Spell Around: Lightweight Alignment Amplification via Rank-One Safety Injection
Harethah Abu Shairah, Hasan Abed Al Kader Hammoud, George Turkiyyah, Bernard Ghanem
https://arxiv.org/abs/2508.20766

Turning the Spell Around: Lightweight Alignment Amplification via Rank-One Safety Injection
Safety alignment in Large Language Models (LLMs) often involves mediating internal representations to refuse harmful requests. Recent research has demonstrated that these safety mechanisms can be bypassed by ablating or removing specific representational directions within the model. In this paper, we propose the opposite approach: Rank-One Safety Injection (ROSI), a white-box method that amplifies a model's safety alignment by permanently steering its activations toward the refusal-mediating su…