Tootfinder

ByDeWay: Boost Your multimodal LLM with DEpth prompting in a Training-Free Way
Rajarshi Roy, Devleena Das, Ankesh Banerjee, Arjya Bhattacharjee, Kousik Dasgupta, Subarna Tripathi
https://arxiv.org/abs/2507.08679

ByDeWay: Boost Your multimodal LLM with DEpth prompting in a Training-Free Way
We introduce ByDeWay, a training-free framework designed to enhance the performance of Multimodal Large Language Models (MLLMs). ByDeWay uses a novel prompting strategy called Layered-Depth-Based Prompting (LDP), which improves spatial reasoning and grounding without modifying any model parameters. It segments the scene into closest, mid-range, and farthest layers using monocular depth estimation, then generates region-specific captions with a grounded vision-language model. These structured, d…

Why is Your Language Model a Poor Implicit Reward Model?
Noam Razin, Yong Lin, Jiarui Yao, Sanjeev Arora
https://arxiv.org/abs/2507.07981 https://

Why is Your Language Model a Poor Implicit Reward Model?
Reward models are key to language model post-training and inference pipelines. Conveniently, recent work showed that every language model defines an implicit reward model (IM-RM), without requiring any architectural changes. However, such IM-RMs tend to generalize worse, especially out-of-distribution, compared to explicit reward models (EX-RMs) that apply a dedicated linear head over the hidden representations of a language model. The existence of a generalization gap is puzzling, as EX-RMs an…

Diffusion Dataset Condensation: Training Your Diffusion Model Faster with Less Data
Rui Huang, Shitong Shao, Zikai Zhou, Pukun Zhao, Hangyu Guo, Tian Ye, Lichen Bai, Shuo Yang, Zeke Xie
https://arxiv.org/abs/2507.05914

Diffusion Dataset Condensation: Training Your Diffusion Model Faster with Less Data
Diffusion models have achieved remarkable success in various generative tasks, but training them remains highly resource-intensive, often requiring millions of images and many days of GPU computation. From a data-centric perspective addressing this limitation, we study diffusion dataset condensation as a new and challenging problem setting. The goal is to construct a "synthetic" sub-dataset with significantly fewer samples than the original dataset, enabling high-quality diffusion model trainin…

EditMF: Drawing an Invisible Fingerprint for Your Large Language Models
Jiaxuan Wu, Yinghan Zhou, Wanli Peng, Yiming Xue, Juan Wen, Ping Zhong
https://arxiv.org/abs/2508.08836 h…

EditMF: Drawing an Invisible Fingerprint for Your Large Language Models
Training large language models (LLMs) is resource-intensive and expensive, making protecting intellectual property (IP) for LLMs crucial. Recently, embedding fingerprints into LLMs has emerged as a prevalent method for establishing model ownership. However, existing back-door-based methods suffer from limited stealth and efficiency. To simultaneously address these issues, we propose EditMF, a training-free fingerprinting paradigm that achieves highly imperceptible fingerprint embedding with min…

Our Revolution here.
It's crunch time in Congress
- Trump's budget, Iran, & oligarch money in politics.
Join the Virtual Lobby Training this Sun 6/22
at 7pm ET (4pm PT).
Our Revolution's DC team is facilitating virtual lobby visits with Capitol Hill offices
and we need your help.
Join and get trained up to speak directly with DC offices via Zoom.
RSVP now
👉

Our Revolution Virtual Lobby Training
Our Revolution Virtual Lobby Training Your elected officials in Congress need to hear from you to build support for progressive policies and make democracy work for the many! Come learn about Our Revolution’s virtual lobby model, how you can get involved, and opportunities to join your neighbors to meet with your Senators and Representatives. Event Details: Date: Sunday, June 22, 2025 Time: 7pm ET (4pm PT) Location: Virtual Training - Zoom Link

⛓️‍💥 Broken Tokens? Your Language Model can Secretly Handle Non-Canonical Tokenizations
#ai #llm

Broken Tokens? Your Language Model can Secretly Handle Non-Canonical Tokenizations
Modern tokenizers employ deterministic algorithms to map text into a single "canonical" token sequence, yet the same string can be encoded as many non-canonical tokenizations using the tokenizer vocabulary. In this work, we investigate the robustness of LMs to text encoded with non-canonical tokenizations entirely unseen during training. Surprisingly, when evaluated across 20 benchmarks, we find that instruction-tuned models retain up to 93.4% of their original performance when given a randomly…

SPACE: Your Genomic Profile Predictor is a Powerful DNA Foundation Model
Zhao Yang, Jiwei Zhu, Bing Su
https://arxiv.org/abs/2506.01833 https://

SPACE: Your Genomic Profile Predictor is a Powerful DNA Foundation Model
Inspired by the success of unsupervised pre-training paradigms, researchers have applied these approaches to DNA pre-training. However, we argue that these approaches alone yield suboptimal results because pure DNA sequences lack sufficient information, since their functions are regulated by genomic profiles like chromatin accessibility. Here, we demonstrate that supervised training for genomic profile prediction serves as a more effective alternative to pure sequence pre-training. Furthermore,…

Your Agent Can Defend Itself against Backdoor Attacks
Li Changjiang, Liang Jiacheng, Cao Bochuan, Chen Jinghui, Wang Ting
https://arxiv.org/abs/2506.08336 …

Your Agent Can Defend Itself against Backdoor Attacks
Despite their growing adoption across domains, large language model (LLM)-powered agents face significant security risks from backdoor attacks during training and fine-tuning. These compromised agents can subsequently be manipulated to execute malicious operations when presented with specific triggers in their inputs or environments. To address this pressing risk, we present ReAgent, a novel defense against a range of backdoor attacks on LLM-based agents. Intuitively, backdoor attacks often res…

This https://arxiv.org/abs/2504.13416 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csLG_…

STAMP Your Content: Proving Dataset Membership via Watermarked Rephrasings
Given how large parts of publicly available text are crawled to pretrain large language models (LLMs), data creators increasingly worry about the inclusion of their proprietary data for model training without attribution or licensing. Their concerns are also shared by benchmark curators whose test-sets might be compromised. In this paper, we present STAMP, a framework for detecting dataset membership-i.e., determining the inclusion of a dataset in the pretraining corpora of LLMs. Given an origi…