Tootfinder

ARMOR: Aligning Secure and Safe Large Language Models via Meticulous Reasoning
Zhengyue Zhao, Yingzi Ma, Somesh Jha, Marco Pavone, Chaowei Xiao
https://arxiv.org/abs/2507.11500

ARMOR: Aligning Secure and Safe Large Language Models via Meticulous Reasoning
Large Language Models (LLMs) have demonstrated remarkable generative capabilities. However, their susceptibility to misuse has raised significant safety concerns. While post-training safety alignment methods have been widely adopted, LLMs remain vulnerable to malicious instructions that can bypass safety constraints. Recent efforts have introduced inference-time safety reasoning (system-2 alignment), where LLMs conduct a reasoning process to perform safety verification before final response. We…

Flocking as a second-order phase transition in self-aligning active crystals
Marco Musacchio, Alexander P. Antonov, Hartmut L\"owen, Lorenzo Caprini
https://arxiv.org/abs/2506.12967

Flocking as a second-order phase transition in self-aligning active crystals
We study a two-dimensional crystal composed of active units governed by self-alignment. This mechanism induces a torque that aligns a particle's orientation with its velocity and leads to a phase transition from a disordered to a flocking crystal. Here, we provide the first microscopic theory that analytically maps the crystal dynamics onto a Landau-Ginzburg model, in which the velocity-dependent effective free energy undergoes a transition from a single-well shape to a Mexican-hat profile. As …

Internal Value Alignment in Large Language Models through Controlled Value Vector Activation
Haoran Jin, Meng Li, Xiting Wang, Zhihao Xu, Minlie Huang, Yantao Jia, Defu Lian
https://arxiv.org/abs/2507.11316

Internal Value Alignment in Large Language Models through Controlled Value Vector Activation
Aligning Large Language Models (LLMs) with human values has attracted increasing attention since it provides clarity, transparency, and the ability to adapt to evolving scenarios. In this paper, we introduce a Controlled Value Vector Activation (ConVA) method that directly aligns the internal values of LLMs by interpreting how a value is encoded in their latent representations and modifies relevant activations to ensure consistent values in LLMs. To ensure an accurate and unbiased interpretatio…

The Space Between Us: A Methodological Framework for Researching Bonding and Proxemics in Situated Group-Agent Interactions
Ana M\"uller, Anja Richert
https://arxiv.org/abs/2506.11829

The Space Between Us: A Methodological Framework for Researching Bonding and Proxemics in Situated Group-Agent Interactions
This paper introduces a multimethod framework for studying spatial and social dynamics in real-world group-agent interactions with socially interactive agents. Drawing on proxemics and bonding theories, the method combines subjective self-reports and objective spatial tracking. Applied in two field studies in a museum (N = 187) with a robot and a virtual agent, the paper addresses the challenges in aligning human perception and behavior. We focus on presenting an open source, scalable, and fiel…

LLM-Driven Dual-Level Multi-Interest Modeling for Recommendation
Ziyan Wang, Yingpeng Du, Zhu Sun, Jieyi Bi, Haoyan Chua, Tianjun Wei, Jie Zhang
https://arxiv.org/abs/2507.10917

LLM-Driven Dual-Level Multi-Interest Modeling for Recommendation
Recently, much effort has been devoted to modeling users' multi-interests based on their behaviors or auxiliary signals. However, existing methods often rely on heuristic assumptions, e.g., co-occurring items indicate the same interest of users, failing to capture user multi-interests aligning with real-world scenarios. While large language models (LLMs) show significant potential for multi-interest analysis due to their extensive knowledge and powerful reasoning capabilities, two key challenge…

""[…] Red Hat Enterprise Linux for Business Developers […] provides self-serve, no-cost access to Red Hat Enterprise Linux [#RHEL] for enterprise development use.""

From Outcomes to Processes: Guiding PRM Learning from ORM for Inference-Time Alignment
Bin Xie, Bingbing Xu, Yige Yuan, Shengmao Zhu, Huawei Shen
https://arxiv.org/abs/2506.12446 …

From Outcomes to Processes: Guiding PRM Learning from ORM for Inference-Time Alignment
Inference-time alignment methods have gained significant attention for their efficiency and effectiveness in aligning large language models (LLMs) with human preferences. However, existing dominant approaches using reward-guided search (RGS) primarily rely on outcome reward models (ORMs), which suffer from a critical granularity mismatch: ORMs are designed to provide outcome rewards for complete responses, while RGS methods rely on process rewards to guide the policy, leading to inconsistent sc…

FMC: Formalization of Natural Language Mathematical Competition Problems
Jiaxuan Xie, Chengwu Liu, Ye Yuan, Siqi Li, Zhiping Xiao, Ming Zhang
https://arxiv.org/abs/2507.11275

FMC: Formalization of Natural Language Mathematical Competition Problems
Efficient and accurate autoformalization methods, which leverage large-scale datasets of extensive natural language mathematical problems to construct formal language datasets, are key to advancing formal mathematical reasoning. In this paper, we propose an autoformalization pipeline based on large language models with error feedback, achieving a fully automatic and training-free formalization approach. Using this pipeline, we curate an Olympiad-level dataset aligning natural language problems …

Aligning Proteins and Language: A Foundation Model for Protein Retrieval
Qifeng Wu, Zhengzhe Liu, Han Zhu, Yizhou Zhao, Daisuke Kihara, Min Xu
https://arxiv.org/abs/2506.08023

Aligning Proteins and Language: A Foundation Model for Protein Retrieval
This paper aims to retrieve proteins with similar structures and semantics from large-scale protein dataset, facilitating the functional interpretation of protein structures derived by structural determination methods like cryo-Electron Microscopy (cryo-EM). Motivated by the recent progress of vision-language models (VLMs), we propose a CLIP-style framework for aligning 3D protein structures with functional annotations using contrastive learning. For model training, we propose a large-scale dat…

Mallows Model with Learned Distance Metrics: Sampling and Maximum Likelihood Estimation
Yeganeh Alimohammadi, Kiana Asgari
https://arxiv.org/abs/2507.08108

Mallows Model with Learned Distance Metrics: Sampling and Maximum Likelihood Estimation
\textit{Mallows model} is a widely-used probabilistic framework for learning from ranking data, with applications ranging from recommendation systems and voting to aligning language models with human preferences~\cite{chen2024mallows, kleinberg2021algorithmic, rafailov2024direct}. Under this model, observed rankings are noisy perturbations of a central ranking $σ$, with likelihood decaying exponentially in distance from $σ$, i.e, $P (π) \propto \exp\big(-β\cdot d(π, σ)\big),$ where $β> 0…

Multi-Task Reward Learning from Human Ratings
Mingkang Wu, Devin White, Evelyn Rose, Vernon Lawhern, Nicholas R Waytowich, Yongcan Cao
https://arxiv.org/abs/2506.09183

Multi-Task Reward Learning from Human Ratings
Reinforcement learning from human feeback (RLHF) has become a key factor in aligning model behavior with users' goals. However, while humans integrate multiple strategies when making decisions, current RLHF approaches often simplify this process by modeling human reasoning through isolated tasks such as classification or regression. In this paper, we propose a novel reinforcement learning (RL) method that mimics human decision-making by jointly considering multiple tasks. Specifically, we lever…

Visual Semantic Description Generation with MLLMs for Image-Text Matching
Junyu Chen, Yihua Gao, Mingyong Li
https://arxiv.org/abs/2507.08590 https://

Visual Semantic Description Generation with MLLMs for Image-Text Matching
Image-text matching (ITM) aims to address the fundamental challenge of aligning visual and textual modalities, which inherently differ in their representations, continuous, high-dimensional image features vs. discrete, structured text. We propose a novel framework that bridges the modality gap by leveraging multimodal large language models (MLLMs) as visual semantic parsers. By generating rich Visual Semantic Descriptions (VSD), MLLMs provide semantic anchor that facilitate cross-modal alignmen…

Stable Preference Optimization for LLMs: A Bilevel Approach Beyond Direct Preference Optimization
Chengtao Jian, Kai Yang, Ye Ouyang, Xiaozhou Ye
https://arxiv.org/abs/2507.07723 …

Stable Preference Optimization for LLMs: A Bilevel Approach Beyond Direct Preference Optimization
Direct Preference Optimization (DPO) has emerged as a popular and efficient alternative to reward modeling and reinforcement learning for aligning language models with human preferences. Despite its empirical success, the theoretical properties and intrinsic limitations of DPO remain underexplored. In this work, we first present a comprehensive analysis of DPO's dynamics from a probability evolution perspective. Our analysis reveals that DPO is highly sensitive to initialization. It also tends …

Trump threatens 10% tariff on countries backing BRICS 'anti-American policy': https://benborges.xyz/2025/07/07/trump-threatens-tariff-on-countries.html

Approaching Dialogue State Tracking via Aligning Speech Encoders and LLMs
\v{S}imon Sedl\'a\v{c}ek, Bolaji Yusuf, J\'an \v{S}vec, Pradyoth Hegde, Santosh Kesiraju, Old\v{r}ich Plchot, Jan \v{C}ernock\'y
https://arxiv.org/abs/2506.08633

Approaching Dialogue State Tracking via Aligning Speech Encoders and LLMs
In this work, we approach spoken Dialogue State Tracking (DST) by bridging the representation spaces of speech encoders and LLMs via a small connector module, with a focus on fully open-sourced and open-data components (WavLM-large, OLMo). We focus on ablating different aspects of such systems including full/LoRA adapter fine-tuning, the effect of agent turns in the dialogue history, as well as fuzzy matching-based output post-processing, which greatly improves performance of our systems on nam…

Joint Optimization-based Targetless Extrinsic Calibration for Multiple LiDARs and GNSS-Aided INS of Ground Vehicles
Junhui Wang, Yan Qiao, Chao Gao, Naiqi Wu
https://arxiv.org/abs/2507.08349

Joint Optimization-based Targetless Extrinsic Calibration for Multiple LiDARs and GNSS-Aided INS of Ground Vehicles
Accurate extrinsic calibration between multiple LiDAR sensors and a GNSS-aided inertial navigation system (GINS) is essential for achieving reliable sensor fusion in intelligent mining environments. Such calibration enables vehicle-road collaboration by aligning perception data from vehicle-mounted sensors to a unified global reference frame. However, existing methods often depend on artificial targets, overlapping fields of view, or precise trajectory estimation, which are assumptions that may…

AR2: Attention-Guided Repair for the Robustness of CNNs Against Common Corruptions
Fuyuan Zhang, Qichen Wang, Jianjun Zhao
https://arxiv.org/abs/2507.06332

AR2: Attention-Guided Repair for the Robustness of CNNs Against Common Corruptions
Deep neural networks suffer from significant performance degradation when exposed to common corruptions such as noise, blur, weather, and digital distortions, limiting their reliability in real-world applications. In this paper, we propose AR2 (Attention-Guided Repair for Robustness), a simple yet effective method to enhance the corruption robustness of pretrained CNNs. AR2 operates by explicitly aligning the class activation maps (CAMs) between clean and corrupted images, encouraging the model…

This https://arxiv.org/abs/2505.07270 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csSE_…

Automated Repair of Ambiguous Natural Language Requirements
The widespread adoption of large language models (LLMs) in software engineering has amplified the role of natural language (NL). The inherent ambiguity of NL threatens software quality, because ambiguous requirements may lead to faulty program generation. The complexity of ambiguity detection and resolution motivates us to introduce automated repair of ambiguous NL requirements, which we approach by reducing code generation uncertainty and aligning NL with input-output examples. Repairing ambig…

Aligning visual imagery to the operator improves geospatial situation awareness in a single-display 360-degree periscope concept https://cognitiveresearchjournal.springeropen.com/articles/10.1186/s41235-025-00646-1

Aligning visual imagery to the operator improves geospatial situation awareness in a single-display 360-degree periscope concept - Cognitive Research: Principles and Implications
Technological advances mean that it is now possible to represent the entire 360° view of the horizon to a submarine periscope operator simultaneously, in strips on a single display, as opposed to the restricted view offered through a conventional periscope aperture. Initial research showing performance improvements for such panoramic displays is promising. However, that research has yet to consider the importance of alignment between the visual representation of the environment on the periscop…

From Threat to Tool: Leveraging Refusal-Aware Injection Attacks for Safety Alignment
Kyubyung Chae, Hyunbin Jin, Taesup Kim
https://arxiv.org/abs/2506.10020

From Threat to Tool: Leveraging Refusal-Aware Injection Attacks for Safety Alignment
Safely aligning large language models (LLMs) often demands extensive human-labeled preference data, a process that's both costly and time-consuming. While synthetic data offers a promising alternative, current methods frequently rely on complex iterative prompting or auxiliary models. To address this, we introduce Refusal-Aware Adaptive Injection (RAAI), a straightforward, training-free, and model-agnostic framework that repurposes LLM attack techniques. RAAI works by detecting internal refusal…

AMRScan: A hybrid R and Nextflow toolkit for rapid antimicrobial resistance gene detection from sequencing data
Kaitao Lai
https://arxiv.org/abs/2507.08062

AMRScan: A hybrid R and Nextflow toolkit for rapid antimicrobial resistance gene detection from sequencing data
AMRScan is a hybrid bioinformatics toolkit implemented in both R and [Nextflow](https://www.nextflow.io/) for the rapid and reproducible detection of antimicrobial resistance (AMR) genes from next-generation sequencing (NGS) data. The toolkit enables users to identify AMR gene hits in sequencing reads by aligning them against reference databases such as CARD using BLAST. The R implementation provides a concise, script-based approach suitable for single-sample analysis, teaching, and rapid pro…

SpecCLIP: Aligning and Translating Spectroscopic Measurements for Stars
Xiaosheng Zhao, Yang Huang, Guirong Xue, Xiao Kong, Jifeng Liu, Xiaoyu Tang, Timothy C. Beers, Yuan-Sen Ting, A-Li Luo
https://arxiv.org/abs/2507.01939

SpecCLIP: Aligning and Translating Spectroscopic Measurements for Stars
In recent years, large language models (LLMs) have transformed natural language understanding through vast datasets and large-scale parameterization. Inspired by this success, we present SpecCLIP, a foundation model framework that extends LLM-inspired methodologies to stellar spectral analysis. Stellar spectra, akin to structured language, encode rich physical and chemical information about stars. By training foundation models on large-scale spectral datasets, our goal is to learn robust and in…

Quantile Reward Policy Optimization: Alignment with Pointwise Regression and Exact Partition Functions
Simon Matrenok, Skander Moalla, Caglar Gulcehre
https://arxiv.org/abs/2507.08068 https://arxiv.org/pdf/2507.08068 https://arxiv.org/html/2507.08068
arXiv:2507.08068v1 Announce Type: new
Abstract: Aligning large language models with pointwise absolute rewards has so far required online, on-policy algorithms such as PPO and GRPO. In contrast, simpler methods that can leverage offline or off-policy data, such as DPO and REBEL, are limited to learning from preference pairs or relative signals. To bridge this gap, we introduce \emph{Quantile Reward Policy Optimization} (QRPO), which learns from pointwise absolute rewards while preserving the simplicity and offline applicability of DPO-like methods. QRPO uses quantile rewards to enable regression to the closed-form solution of the KL-regularized RL objective. This reward yields an analytically tractable partition function, removing the need for relative signals to cancel this term. Moreover, QRPO scales with increased compute to estimate quantile rewards, opening a new dimension for pre-computation scaling. Empirically, QRPO consistently achieves top performance on chat and coding evaluations -- reward model scores, AlpacaEval 2, and LeetCode -- compared to DPO, REBEL, and SimPO across diverse datasets and 8B-scale models. Finally, we find that training with robust rewards instead of converting them to preferences induces less length bias.
toXiv_bot_toot

This is the next stages of Columbus’s long-running effort to update its 1950s-era zoning code. This phase will incorporate a housing component to address affordability issues by better aligning work and residential areas.
We also need to build more houses.
https://

ChainEdit: Propagating Ripple Effects in LLM Knowledge Editing through Logical Rule-Guided Chains
Zilu Dong, Xiangqing Shen, Zinong Yang, Rui Xia
https://arxiv.org/abs/2507.08427 …

ChainEdit: Propagating Ripple Effects in LLM Knowledge Editing through Logical Rule-Guided Chains
Current knowledge editing methods for large language models (LLMs) struggle to maintain logical consistency when propagating ripple effects to associated facts. We propose ChainEdit, a framework that synergizes knowledge graph-derived logical rules with LLM logical reasoning capabilities to enable systematic chain updates. By automatically extracting logical patterns from structured knowledge bases and aligning them with LLMs' internal logics, ChainEdit dynamically generates and edits logically…

Q&A with Hugging Face Chief Ethics Scientist Margaret Mitchell on aligning AI development with human needs, the "illusion of consensus" around AGI, and more (Melissa Heikkilä/Financial Times)
https://www.ft.com/content/7089bff2-25fc-4a25-98bf-8828ab24…

Margaret Mitchell: artificial general intelligence is ‘just vibes and snake oil’
One of the pioneers of AI ethics explains why human needs should be the central driver in the development of the technology

CoMemo: LVLMs Need Image Context with Image Memory
Shi Liu, Weijie Su, Xizhou Zhu, Wenhai Wang, Jifeng Dai
https://arxiv.org/abs/2506.06279 https://…

CoMemo: LVLMs Need Image Context with Image Memory
Recent advancements in Large Vision-Language Models built upon Large Language Models have established aligning visual features with LLM representations as the dominant paradigm. However, inherited LLM architectural designs introduce suboptimal characteristics for multimodal processing. First, LVLMs exhibit a bimodal distribution in attention allocation, leading to the progressive neglect of middle visual content as context expands. Second, conventional positional encoding schemes fail to preser…

# Philosophical test fails ChatGPT: AI coherence isn’t enough to prove human mind
The research reveals that #ChatGPT does exhibit proficiency in basic coherence building. It maintains consistent dictional and intentional lines by reusing phrases and aligning responses with contextual topics. It also demonstrates some ability to construct rational coherence by offering logically consistent replies…

Leveraging machine learning features for linear optical interferometer control
Sergei S. Kuzmin, Ivan V. Dyakonov, Stanislav S. Straupe
https://arxiv.org/abs/2505.24032

Leveraging machine learning features for linear optical interferometer control
We have developed an algorithm that constructs a model of a reconfigurable optical interferometer, independent of specific architectural constraints. The programming of unitary transformations on the interferometer's optical modes relies on either an analytical method for deriving the unitary matrix from a set of phase shifts or an optimization routine when such decomposition is not available. Our algorithm employs a supervised learning approach, aligning the interferometer model with a trainin…

Stochastically Dominant Peer Prediction
Yichi Zhang, Shengwei Xu, David Pennock, Grant Schoenebeck
https://arxiv.org/abs/2506.02259 https://

Stochastically Dominant Peer Prediction
Eliciting reliable human feedback is essential for many machine learning tasks, such as learning from noisy labels and aligning AI systems with human preferences. Peer prediction mechanisms incentivize truthful reporting without ground truth verification by scoring agents based on correlations with peers. Traditional mechanisms, which ensure that truth-telling maximizes the expected scores in equilibrium, can elicit honest information while assuming agents' utilities are linear functions of the…

This https://arxiv.org/abs/2503.07217 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csSD_…

ReelWave: Multi-Agentic Movie Sound Generation through Multimodal LLM Conversation
Current audio generation conditioned by text or video focuses on aligning audio with text/video modalities. Despite excellent alignment results, these multimodal frameworks still cannot be directly applied to compelling movie storytelling involving multiple scenes, where "on-screen" sounds require temporally-aligned audio generation, while "off-screen" sounds contribute to appropriate environment sounds accompanied by background music when applicable. Inspired by professional movie production, …

ROS-related Robotic Systems Development with V-model-based Application of MeROS Metamodel
Tomasz Winiarski, Jan Kaniuka, Daniel Gie{\l}dowski, Jakub Ostrysz, Krystian Radlak, Dmytro Kushnir
https://arxiv.org/abs/2506.08706

ROS-related Robotic Systems Development with V-model-based Application of MeROS Metamodel
As robotic systems grow increasingly complex, heterogeneous, and safety-critical, the need for structured development methodologies becomes paramount. Although frameworks like the Robot Operating System (ROS) and Model-Based Systems Engineering (MBSE) offer foundational tools, they often lack integration when used together. This paper addresses that gap by aligning the widely recognized V-model development paradigm with the MeROS metamodel SysML-based modeling language tailored for ROS-based sy…

KERAG_R: Knowledge-Enhanced Retrieval-Augmented Generation for Recommendation
Zeyuan Meng, Zixuan Yi, Iadh Ounis
https://arxiv.org/abs/2507.05863 https://

KERAG_R: Knowledge-Enhanced Retrieval-Augmented Generation for Recommendation
Large Language Models (LLMs) have shown strong potential in recommender systems due to their contextual learning and generalisation capabilities. Existing LLM-based recommendation approaches typically formulate the recommendation task using specialised prompts designed to leverage their contextual abilities, and aligning their outputs closely with human preferences to yield an improved recommendation performance. However, the use of LLMs for recommendation tasks is limited by the absence of dom…

Aligning Protein Conformation Ensemble Generation with Physical Feedback
Jiarui Lu, Xiaoyin Chen, Stephen Zhewen Lu, Aur\'elie Lozano, Vijil Chenthamarakshan, Payel Das, Jian Tang
https://arxiv.org/abs/2505.24203

Aligning Protein Conformation Ensemble Generation with Physical Feedback
Protein dynamics play a crucial role in protein biological functions and properties, and their traditional study typically relies on time-consuming molecular dynamics (MD) simulations conducted in silico. Recent advances in generative modeling, particularly denoising diffusion models, have enabled efficient accurate protein structure prediction and conformation sampling by learning distributions over crystallographic structures. However, effectively integrating physical supervision into these d…

This https://arxiv.org/abs/2402.17732 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_mat…

Batched Nonparametric Contextual Bandits
We study nonparametric contextual bandits under batch constraints, where the expected reward for each action is modeled as a smooth function of covariates, and the policy updates are made at the end of each batch of observations. We establish a minimax regret lower bound for this setting and propose a novel batch learning algorithm that achieves the optimal regret (up to logarithmic factors). In essence, our procedure dynamically splits the covariate space into smaller bins, carefully aligning …

This https://arxiv.org/abs/2501.07071 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csAI_…

Value Compass Leaderboard: A Platform for Fundamental and Validated Evaluation of LLMs Values
As Large Language Models (LLMs) achieve remarkable breakthroughs, aligning their values with humans has become imperative for their responsible development and customized applications. However, there still lack evaluations of LLMs values that fulfill three desirable goals. (1) Value Clarification: We expect to clarify the underlying values of LLMs precisely and comprehensively, while current evaluations focus narrowly on safety risks such as bias and toxicity. (2) Evaluation Validity: Existing …

HYFuse: Aligning Heterogeneous Speech Pre-Trained Representations in Hyperbolic Space for Speech Emotion Recognition
Orchid Chetia Phukan, Girish, Mohd Mujtaba Akhtar, Swarup Ranjan Behera, Pailla Balakrishna Reddy, Arun Balaji Buduru, Rajesh Sharma
https://arxiv.org/abs/2506.03403

HYFuse: Aligning Heterogeneous Speech Pre-Trained Representations in Hyperbolic Space for Speech Emotion Recognition
Compression-based representations (CBRs) from neural audio codecs such as EnCodec capture intricate acoustic features like pitch and timbre, while representation-learning-based representations (RLRs) from pre-trained models trained for speech representation learning such as WavLM encode high-level semantic and prosodic information. Previous research on Speech Emotion Recognition (SER) has explored both, however, fusion of CBRs and RLRs haven't been explored yet. In this study, we solve this gap…

This https://arxiv.org/abs/2410.05605 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csSE_…

CodeDPO: Aligning Code Models with Self Generated and Verified Source Code
Code generation models have shown significant potential for programming tasks. However, existing training methods like supervised fine-tuning face key limitations: they do not effectively teach models to prioritize correct over incorrect solutions in ambiguous situations, nor do they effectively optimize the runtime efficiency of the generated code. To address these challenges, we propose CodeDPO, a framework that integrates preference learning into code generation to improve two key code prefe…

Offline Map Matching Based on Localization Error Distribution Modeling
Ruilin Xu, Yuchen Song, Kaijie Li, Xitong Gao, Kejiang Ye, Fan Zhang, Juanjuan Zhao
https://arxiv.org/abs/2505.23123

Offline Map Matching Based on Localization Error Distribution Modeling
Offline map matching involves aligning historical trajectories of mobile objects, which may have positional errors, with digital maps. This is essential for applications in intelligent transportation systems (ITS), such as route analysis and traffic pattern mining. Existing methods have two main limitations: (i) they assume a uniform Localization Error Distribution (LED) across urban areas, neglecting environmental factors that lead to suboptimal path search ranges, and (ii) they struggle to ef…

I got the tow hitch installed on the Crosstrek today. Holding a tow hitch bar up, while aligning a bolt, and trying to get the nut on said bolt before I lose the bolt inside the frame OR drop the hitch on my face can be quite exciting. Then torquing those to 110 ft/lbs while lying under all of that can be hard on your shoulder. I'm exhausted now.

MPF: Aligning and Debiasing Language Models post Deployment via Multi Perspective Fusion
Xin Guan, PeiHsin Lin, Zekun Wu, Ze Wang, Ruibo Zhang, Emre Kazim, Adriano Koshiyama
https://arxiv.org/abs/2507.02595

MPF: Aligning and Debiasing Language Models post Deployment via Multi Perspective Fusion
Multiperspective Fusion (MPF) is a novel posttraining alignment framework for large language models (LLMs) developed in response to the growing need for easy bias mitigation. Built on top of the SAGED pipeline, an automated system for constructing bias benchmarks and extracting interpretable baseline distributions, MPF leverages multiperspective generations to expose and align biases in LLM outputs with nuanced, humanlike baselines. By decomposing baseline, such as sentiment distributions from …

Insights from Educators on Building a More Cohesive Quantum Information Science and Engineering Education Ecosystem
Shams El-Adawy, A. R. Pi\~na, Benjamin M. Zwickl, H. J. Lewandowski
https://arxiv.org/abs/2507.01578

Insights from Educators on Building a More Cohesive Quantum Information Science and Engineering Education Ecosystem
As the need for a quantum-ready workforce grows, educators in Quantum Information Science and Engineering (QISE) face the challenge of aligning their programs and courses with industry needs. Through a series of interviews with program directors and faculty across 15 different institutions, we identified the considerations that educators are currently addressing as they develop their various courses and programs. Grounded in a curriculum framework, we conducted a Strengths, Weaknesses, Opportun…

Deformable Medical Image Registration with Effective Anatomical Structure Representation and Divide-and-Conquer Network
Xinke Ma, Yongsheng Pan, Qingjie Zeng, Mengkang Lu, Bolysbek Murat Yerzhanuly, Bazargul Matkerim, Yong Xia
https://arxiv.org/abs/2506.19222

Deformable Medical Image Registration with Effective Anatomical Structure Representation and Divide-and-Conquer Network
Effective representation of Regions of Interest (ROI) and independent alignment of these ROIs can significantly enhance the performance of deformable medical image registration (DMIR). However, current learning-based DMIR methods have limitations. Unsupervised techniques disregard ROI representation and proceed directly with aligning pairs of images, while weakly-supervised methods heavily depend on label constraints to facilitate registration. To address these issues, we introduce a novel ROI-…

Fast entropy-regularized SDP relaxations for permutation synchronization
Michael Lindsey, Yunpeng Shi
https://arxiv.org/abs/2506.20191 https://

Fast entropy-regularized SDP relaxations for permutation synchronization
We introduce fast randomized algorithms for solving semidefinite programming (SDP) relaxations of the partial permutation synchronization (PPS) problem, a core task in multi-image matching with significant relevance to 3D reconstruction. Our methods build on recent advances in entropy-regularized semidefinite programming and are tailored to the unique structure of PPS, in which the unknowns are partial permutation matrices aligning sparse and noisy pairwise correspondences across images. We pro…

CTR-Guided Generative Query Suggestion in Conversational Search
Erxue Min, Hsiu-Yuan Huang, Xihong Yang, Min Yang, Xin Jia, Yunfang Wu, Hengyi Cai, Junfeng Wang, Shuaiqiang Wang, Dawei Yin
https://arxiv.org/abs/2507.04072

CTR-Guided Generative Query Suggestion in Conversational Search
Generating effective query suggestions in conversational search requires aligning model outputs with user preferences, which is challenging due to sparse and noisy click signals. We propose GQS, a generative framework that integrates click modeling and preference optimization to enhance real-world user engagement. GQS consists of three key components: (1) a Multi-Source CTR Modeling module that captures diverse contextual signals to estimate fine-grained click-through rates; (2) a Diversity-Awa…

Case Study for Developing a UXR Point of View for FinOps Product Innovation
Jason Dong, Anna Wu
https://arxiv.org/abs/2506.15314 https://

Case Study for Developing a UXR Point of View for FinOps Product Innovation
In the dynamic landscape of Cloud financial management, we are sharing a case study exploring the development of a User Experience Research (UXR) Point of View (PoV) to drive FinOps product innovation. We demonstrate how qualitative and quantitative research methods working together to navigate the challenges of understanding customer needs, aligning cross-functional teams, and prioritizing limited resources. Through a multi-phased research approach, the research team identifies opportunities, …

Gradient-Adaptive Policy Optimization: Towards Multi-Objective Alignment of Large Language Models
Chengao Li, Hanyu Zhang, Yunkun Xu, Hongyan Xue, Xiang Ao, Qing He
https://arxiv.org/abs/2507.01915

Gradient-Adaptive Policy Optimization: Towards Multi-Objective Alignment of Large Language Models
Reinforcement Learning from Human Feedback (RLHF) has emerged as a powerful technique for aligning large language models (LLMs) with human preferences. However, effectively aligning LLMs with diverse human preferences remains a significant challenge, particularly when they are conflict. To address this issue, we frame human value alignment as a multi-objective optimization problem, aiming to maximize a set of potentially conflicting objectives. We introduce Gradient-Adaptive Policy Optimization…

Optimal alignment of Lorentz orientation and generalization to matrix Lie groups
Congzhou M Sha
https://arxiv.org/abs/2506.14994 https://

Optimal alignment of Lorentz orientation and generalization to matrix Lie groups
There exist elegant methods of aligning point clouds in $\mathbb R^3$. Unfortunately, these methods rely on the positive definite property of the Euclidean metric, and do not easily extend to the indefinite Minkowski metric. In this paper, we propose two solutions to the following problem: given inertial reference frames $A$ and $B$, and given (possibly noisy) measurements of a set of 4-vectors $\{v_i\}$ made in those reference frames with components $\{v_{A,i}\}$ and $\{v_{B,i}\}$, find the op…

BulletGen: Improving 4D Reconstruction with Bullet-Time Generation
Denys Rozumnyi, Jonathon Luiten, Numair Khan, Johannes Sch\"onberger, Peter Kontschieder
https://arxiv.org/abs/2506.18601

BulletGen: Improving 4D Reconstruction with Bullet-Time Generation
Transforming casually captured, monocular videos into fully immersive dynamic experiences is a highly ill-posed task, and comes with significant challenges, e.g., reconstructing unseen regions, and dealing with the ambiguity in monocular depth estimation. In this work we introduce BulletGen, an approach that takes advantage of generative models to correct errors and complete missing information in a Gaussian-based dynamic scene representation. This is done by aligning the output of a diffusion-…

This https://arxiv.org/abs/2505.10640 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csSE_…

The Hitchhikers Guide to Production-ready Trustworthy Foundation Model powered Software (FMware)
Foundation Models (FMs) such as Large Language Models (LLMs) are reshaping the software industry by enabling FMware, systems that integrate these FMs as core components. In this KDD 2025 tutorial, we present a comprehensive exploration of FMware that combines a curated catalogue of challenges with real-world production concerns. We first discuss the state of research and practice in building FMware. We further examine the difficulties in selecting suitable models, aligning high-quality domain-s…

SafePTR: Token-Level Jailbreak Defense in Multimodal LLMs via Prune-then-Restore Mechanism
Beitao Chen, Xinyu Lyu, Lianli Gao, Jingkuan Song, Heng Tao Shen
https://arxiv.org/abs/2507.01513

SafePTR: Token-Level Jailbreak Defense in Multimodal LLMs via Prune-then-Restore Mechanism
By incorporating visual inputs, Multimodal Large Language Models (MLLMs) extend LLMs to support visual reasoning. However, this integration also introduces new vulnerabilities, making MLLMs susceptible to multimodal jailbreak attacks and hindering their safe deployment.Existing defense methods, including Image-to-Text Translation, Safe Prompting, and Multimodal Safety Tuning, attempt to address this by aligning multimodal inputs with LLMs' built-in safeguards.Yet, they fall short in uncovering …

Hybrid-Sep: Language-queried audio source separation via pre-trained Model Fusion and Adversarial Diffusion Training
Jianyuan Feng, Guangzheng Li, Yangfei Xu
https://arxiv.org/abs/2506.16833

Hybrid-Sep: Language-queried audio source separation via pre-trained Model Fusion and Adversarial Diffusion Training
Language-queried Audio Separation (LASS) employs linguistic queries to isolate target sounds based on semantic descriptions. However, existing methods face challenges in aligning complex auditory features with linguistic context while preserving separation precision. Current research efforts focus primarily on text description augmentation and architectural innovations, yet the potential of integrating pre-trained self-supervised learning (SSL) audio models and Contrastive Language-Audio Pretra…

Robust Alignment via Partial Gromov-Wasserstein Distances
Xiaoyun Gong, Sloan Nietert, Ziv Goldfeld
https://arxiv.org/abs/2506.21507 https://

Robust Alignment via Partial Gromov-Wasserstein Distances
The Gromov-Wasserstein (GW) problem provides a powerful framework for aligning heterogeneous datasets by matching their internal structures in a way that minimizes distortion. However, GW alignment is sensitive to data contamination by outliers, which can greatly distort the resulting matching scheme. To address this issue, we study robust GW alignment, where upon observing contaminated versions of the clean data distributions, our goal is to accurately estimate the GW alignment cost between th…

AgentDistill: Training-Free Agent Distillation with Generalizable MCP Boxes
Jiahao Qiu, Xinzhe Juan, Yimin Wang, Ling Yang, Xuan Qi, Tongcheng Zhang, Jiacheng Guo, Yifu Lu, Zixin Yao, Hongru Wang, Shilong Liu, Xun Jiang, Liu Leqi, Mengdi Wang
https://arxiv.org/abs/2506.14728

AgentDistill: Training-Free Agent Distillation with Generalizable MCP Boxes
While knowledge distillation has become a mature field for compressing large language models (LLMs) into smaller ones by aligning their outputs or internal representations, the distillation of LLM-based agents, which involve planning, memory, and tool use, remains relatively underexplored. Existing agent distillation methods typically replay full teacher trajectories or imitate step-by-step teacher tool usage, but they often struggle to train student agents to dynamically plan and act in novel …

Replaced article(s) found for cs.CL. https://arxiv.org/list/cs.CL/new
[3/3]:
- Aligning Frozen LLMs by Reinforcement Learning: An Iterative Reweight-then-Optimize Approach
Zhang, Li, Zeng, Li, Wang, Lin, Lu, Garcia, Hong

SAFER: Probing Safety in Reward Models with Sparse Autoencoder
Sihang Li, Wei Shi, Ziyuan Xie, Tao Liang, Guojun Ma, Xiang Wang
https://arxiv.org/abs/2507.00665

SAFER: Probing Safety in Reward Models with Sparse Autoencoder
Reinforcement learning from human feedback (RLHF) is a key paradigm for aligning large language models (LLMs) with human values, yet the reward models at its core remain largely opaque. In this work, we present sparse Autoencoder For Enhanced Reward model (\textbf{SAFER}), a novel framework for interpreting and improving reward models through mechanistic analysis. Leveraging Sparse Autoencoders (SAEs), we uncover human-interpretable features in reward model activations, enabling insight into sa…

CoVE: Compressed Vocabulary Expansion Makes Better LLM-based Recommender Systems
Haochen Zhang, Tianyi Zhang, Junze Yin, Oren Gal, Anshumali Shrivastava, Vladimir Braverman
https://arxiv.org/abs/2506.19993

CoVE: Compressed Vocabulary Expansion Makes Better LLM-based Recommender Systems
Recommender systems play a pivotal role in providing relevant content to users. With the rapid development of large language models (LLMs), researchers have begun utilizing LLMs to build more powerful recommender systems. However, existing approaches that focus on aligning LLMs with recommendation tasks do not fully leverage their sequential information processing capabilities, leading to suboptimal performance. In this paper, we propose a novel system called compressed vocabulary expansion (…

Not All Jokes Land: Evaluating Large Language Models Understanding of Workplace Humor
Moahmmadamin Shafiei, Hamidreza Saffari
https://arxiv.org/abs/2506.01819

Not All Jokes Land: Evaluating Large Language Models Understanding of Workplace Humor
With the recent advances in Artificial Intelligence (AI) and Large Language Models (LLMs), the automation of daily tasks, like automatic writing, is getting more and more attention. Hence, efforts have focused on aligning LLMs with human values, yet humor, particularly professional industrial humor used in workplaces, has been largely neglected. To address this, we develop a dataset of professional humor statements along with features that determine the appropriateness of each statement. Our ev…

Defining the Game Producer: A Mapping of Key Characteristics and Differentiators of the Professional Behind Digital Game Production
Rafael C. Lopes, Danilo M. Ribeiro
https://arxiv.org/abs/2506.14409

Defining the Game Producer: A Mapping of Key Characteristics and Differentiators of the Professional Behind Digital Game Production
Introduction: As digital games grow in complexity, the role of the Game Producer becomes increasingly relevant for aligning creative, technical, and business dimensions. Objective: This study aimed to identify and map the main characteristics, skills, and competencies that define the Digital Game Producer profile. Methodology: A qualitative investigation was conducted with 11 semi-structured interviews, analyzed through Grounded Theory to build categories grounded in professional practice. Resu…

Massive Supervised Fine-tuning Experiments Reveal How Data, Layer, and Training Factors Shape LLM Alignment Quality
Yuto Harada, Yusuke Yamauchi, Yusuke Oda, Yohei Oseki, Yusuke Miyao, Yu Takagi
https://arxiv.org/abs/2506.14681

Massive Supervised Fine-tuning Experiments Reveal How Data, Layer, and Training Factors Shape LLM Alignment Quality
Supervised fine-tuning (SFT) is a critical step in aligning large language models (LLMs) with human instructions and values, yet many aspects of SFT remain poorly understood. We trained a wide range of base models on a variety of datasets including code generation, mathematical reasoning, and general-domain tasks, resulting in 1,000+ SFT models under controlled conditions. We then identified the dataset properties that matter most and examined the layer-wise modifications introduced by SFT. Our…