All entries

Uncovering and Mitigating Aggregation-Induced Reward Hacking in Multi-Reward Reinforcement Learning

arXiv:2609.00213v1 Announce Type: new Abstract: Reinforcement learning fine-tuning of large language models increasingly adopts multiple reward dimensions, including verifiable rules, task-specific ev

safetyarxiv-cs-cl

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Model Releases

Uncovering the Computational Ingredients of Human-Like Representations in LLMs

DGX agent

arXiv:2510.01030v2 Announce Type: replace Abstract: The human ability to translate diverse perceptual and linguistic inputs into structured behavior has been thought to rest on learning robust represe

model-releasesarxiv-cs-ai

2 Sept 2026

Uncovering Understanding-Generation Synergy in Native Unified Multimodal Models: From Representation, Task to System

arXiv:2609.01607v1 Announce Type: new Abstract: While unified multimodal models (UMMs) jointly perform visual understanding and generation within a single model, functional unification does not guaran

safetyarxiv-cs-cv

Uniform a priori bounds and error analysis for the Adam stochastic gradient descent optimization method

arXiv:2603.18899v2 Announce Type: replace Abstract: The adaptive moment estimation (Adam) optimizer proposed by Kingma & Ba (2014) is presumably the most popular stochastic gradient descent (SGD) opti

Tutorials

Unmasking Face Embeddings: Reading, Rendering and Naming with Foundation Models

arXiv:2609.00411v1 Announce Type: new Abstract: Modern face recognition (FR) owes much of its success to deep neural networks that learn to extract compact identity embeddings from face images. These

tutorialsarxiv-cs-cv

Applications

User Representation via Cross Multi-source Behavior Pre-training for Mobile Games

arXiv:2609.01057v1 Announce Type: new Abstract: User representation pre-training has become a fundamental paradigm for alleviating data sparsity in downstream personalization tasks. However, existing

applicationsarxiv-cs-ai

V-Co: A Closer Look at Visual Representation Alignment via Co-Denoising

arXiv:2603.16792v2 Announce Type: replace-cross Abstract: Pixel-space diffusion has recently re-emerged as a strong alternative to latent diffusion, enabling high-quality generation without pretrained

Validity-Aware Jailbreak Evaluation for Large Language Models

arXiv:2609.00498v1 Announce Type: new Abstract: Jailbreak robustness has become central to large language model (LLM) safety evaluation, yet prevailing methodologies rely primarily on refusal behavior

Value Over Language Model: Detecting Original Contribution in Writing

arXiv:2609.00700v1 Announce Type: new Abstract: LLMs have been rapidly adopted across writing tasks, prompting the development of tools for detecting LLM-generated text. Yet, these tools largely measu

ValueGraph: Value-Signal Guided Graph Pre-training for Contextualized User Representation

arXiv:2609.00057v1 Announce Type: cross Abstract: Value signals are aggregated user-level moral representations that capture users' inferred value-related tendencies from their online discourse. User

Variable Selection for Feature-Based Newsvendor

arXiv:2609.01544v1 Announce Type: cross Abstract: Feature-based newsvendor models use observable covariates to tailor inventory decisions, aiming to balance holding and shortage costs under demand unc

safetyarxiv-cs-lg

Variance-Adaptive Muon: Pre-Orthogonalization Variance Modulation for Efficient Language Model Pretraining

model-releasesarxiv-cs-lg

arXiv:2601.14603v2 Announce Type: replace Abstract: Optimizer design plays a central role in efficient language model pretraining, directly affecting optimization dynamics, convergence speed, and comp

Local Ai

VATO: A Vortex-Force-Aware Transformer Operator for Unsteady Separated Aerofoil Flows

arXiv:2609.00507v1 Announce Type: new Abstract: Accurate prediction of unsteady separated flows is challenging because the aerodynamic loads depend on nonlinear separation and vortex-shedding dynamics

local-aiarxiv-cs-lg

VectorGym: A Multi-Task Benchmark for SVG Code Generation, Sketching and Editing

arXiv:2603.29852v2 Announce Type: replace-cross Abstract: We introduce VectorGym, a comprehensive benchmark suite for Scalable Vector Graphics (SVG) that spans generation from text and sketches, compl

Verdict Instability of OOD Scores under Reference Resampling

arXiv:2609.00691v1 Announce Type: new Abstract: Post-hoc out-of-distribution detectors are fitted on a finite reference set, so every score they produce is an estimate. If we had chosen a different se

Verifiable Disaster Storylines and Causal Knowledge Graphs: A Citation-Grounded Pipeline from Heterogeneous Humanitarian Sources

arXiv:2609.00858v1 Announce Type: new Abstract: Effective humanitarian response depends on the rapid synthesis of heterogeneous, high-volume information sources - a task that routinely exceeds human a

VerNav: Verifier-First Low-Latency Vision-and-Language Navigation

model-releasesarxiv-cs-cv

arXiv:2609.00920v1 Announce Type: cross Abstract: Vision-and-Language Navigation (VLN) requires an agent to navigate through unseen 3D environments according to natural-language instructions. Explicit

VerTox: Verifiable Reward-Guided Corpus Poisoning Against Neural Ranking Models

arXiv:2609.01325v1 Announce Type: new Abstract: Neural ranking models have become core components of modern information retrieval systems and important building blocks of AI systems such as retrieval-

researcharxiv-cs-cl

VIBE-Bench: Evaluating Personalized Large Language Models When Profiles Don't Mean Preferences

arXiv:2609.00921v1 Announce Type: new Abstract: Personalized Large Language Models (PLLMs) aim to tailor responses to individual users, where a central challenge is preference reasoning: inferring que

ViPlan: A Benchmark for Visual Planning with Symbolic Predicates and Vision-Language Models

arXiv:2505.13180v3 Announce Type: replace Abstract: Integrating Large Language Models with symbolic planners is a promising direction for obtaining verifiable and grounded plans, with recent works ext

Local Ai

Vision-Based Leader-Follower Formation Control for Cooperative UAVs in GPS-Degraded Environments

arXiv:2609.01420v1 Announce Type: new Abstract: Cooperation in multi-UAV systems requires reliable relative perception so that follower vehicles can maintain formation and continue their mission safel

local-aiarxiv-cs-ro

Applications

Vision Is Not Overhead: One-Pass Block Drafting for Lossless Speculative Decoding in Vision-Language Models

arXiv:2609.00355v1 Announce Type: new Abstract: Speculative decoding accelerates generation without changing its output, yet on vision-language models (VLMs) it has been caught in a self-defeating cyc

applicationsarxiv-cs-ai

Agents

Vision-Language-Guided Pseudo-Labels for Unsupervised Domain Adaptation in Semantic Segmentation for Waste Sorting

arXiv:2609.00898v1 Announce Type: cross Abstract: Obtaining labeled data for semantic segmentation in applied settings (e.g., autonomous driving, industrial waste sorting) is expensive and often infea

agentsarxiv-cs-ai

Local Ai

Visual Attention Faithfulness in Vision-Language Models is Heterogeneous

arXiv:2609.00830v1 Announce Type: cross Abstract: Whether attention weights faithfully reflect model reasoning has been actively debated in NLP, yet this question remains largely unexplored for the vi

local-aiarxiv-cs-ai

Visual Framing for News Stance Detection via Image Generation

arXiv:2609.00685v1 Announce Type: cross Abstract: Article-level news stance detection aims to identify the perspective of news articles toward social issues. Despite advances in stance detection and i

ViTAL-X: Video-Text Alignment with Cross-Modal Temporal Edits

model-releasesarxiv-cs-cv

arXiv:2609.00505v1 Announce Type: new Abstract: Video-text models adapted from image-text architectures (e.g., CLIP) frequently exhibit temporal blindness, the inability to perceive fundamental cues l

ViTAMINS: An Empirical Study of Training Self-Supervised Vision Transformers with Synthetic Hard Negatives

arXiv:2609.01041v1 Announce Type: cross Abstract: We introduce ViTAMINS, a method that integrates synthetic hard negatives into unsupervised vision transformer pretraining to improve representation qu

Tutorials

(V)LMs generalize beyond surface co-occurrence: Evidence from cross-modal number agreement

arXiv:2609.00443v1 Announce Type: cross Abstract: Language models learn about grammatical number primarily from co-occurrence, and show frequency effects as a result---sometimes taken to indicate that

tutorialsarxiv-cs-ai

VoiceLongMemEval: Do Assistants Remember How You Sounded?

arXiv:2609.00570v1 Announce Type: new Abstract: With the growing scale of multi-agent architectures and large language models, deployed AI assistants are increasingly tasked with reasoning over long,

VOIM: Training-Free Open-Vocabulary 3D Instance Mapping for RGB-D and Monocular SLAM

arXiv:2609.00775v1 Announce Type: cross Abstract: We present Voxel-Grounded Online Instance Manager (VOIM), a training-free voxel-grounded instance manager that builds open-vocabulary 3D instance maps

Wave Function Backpropagation with Explicit Temporal-Interval Dynamics

arXiv:2609.00503v1 Announce Type: new Abstract: Conventional neural networks learn predominantly through affine transformations followed by nonlinear activations, while elapsed time is often treated a

Web Price Extraction: State of the Art and an Adaptive Browserless Implementation

arXiv:2609.01030v1 Announce Type: cross Abstract: Price extraction from websites is a key task for market monitoring, price comparison, and business analytics in e-commerce. Existing approaches can be

Agents

WHALE: A Simple Recipe for Joint Harness-Weight Optimization

arXiv:2609.00196v1 Announce Type: cross Abstract: Agent performance depends jointly on the model parameters and the executable harness code that manages context and control flow. Optimizing either com

agentsarxiv-cs-ai

What Does an Agentic Software Engineering Benchmark Measure? Profiling Task Demands and Agent Behaviour Beyond What Category Labels Reveal

model-releasesarxiv-cs-cl

arXiv:2609.01271v1 Announce Type: cross Abstract: Agentic software engineering benchmarks are typically summarized by nominal category labels such as 'bug fix' or 'feature implementation,' yet benchma

Tutorials

What, Where, and How: Probing Spatiotemporal Representations in Video Foundation Models

arXiv:2609.01551v1 Announce Type: new Abstract: Self-supervised video foundation models learn rich spatiotemporal representations, yet it remains unclear what visual concepts these representations enc

tutorialsarxiv-cs-cv

When Compression Helps and When It Hurts: Condition-Aware Analysis of Chain-of-Thought Distillation

arXiv:2606.21704v2 Announce Type: replace Abstract: Chain-of-Thought (CoT) distillation transfers multi-step reasoning from large reasoning models to smaller students, but verbose teacher traces infla

researcharxiv-cs-cl

When Does Online Adaptation Pay on the Edge? A Leakage-Free Evaluation of Warmup, Learning-Rate Selection, and Resource Trade-offs for Time-Series Forecasting

model-releasesarxiv-cs-lg

arXiv:2609.01126v1 Announce Type: new Abstract: Online adaptation can help edge time-series forecasting under distribution drift, but its measured benefit is sensitive to evaluation choices. We study

When Features Become Instances: Inverted Contrastive Learning for Unsupervised Feature Selection

arXiv:2609.00782v1 Announce Type: new Abstract: Unsupervised feature selection seeks a compact subset of informative features without access to class labels, making feature utility difficult to define

When Guardrails Look Effective: Construct Validity Failures in LLM Agent Commerce Evaluation

arXiv:2609.01519v1 Announce Type: new Abstract: Interactive simulations increasingly evaluate policies in markets populated by language-model agents. Their outputs can look economic---prices, profits,

When Metropolis and Hastings Meet Bradley and Terry: Exact MCMC From Preference Voting

arXiv:2609.00905v1 Announce Type: new Abstract: Sampling from distributions conditioned on desired semantic properties is an emerging challenge in modern generative modeling. Metropolis-Hastings (MH)

When Modality Gap Reduction Fails: Prediction-Level Hubness in CLIP

arXiv:2609.01103v1 Announce Type: new Abstract: Reducing the modality gap between image and text representations in CLIP is widely expected to improve cross-modal alignment and downstream performance.

safetyarxiv-cs-cl

When Prediction Error Is Not Enough: Evaluating Nuisance-Function Prediction for Causal Estimation

arXiv:2609.00071v1 Announce Type: new Abstract: Prediction error is widely used to evaluate nuisance-function estimators in causal inference, but its relationship with causal estimator performance may

When Safety Routing Breaks: Understanding Alignment Fragility under Benign Fine-Tuning

arXiv:2609.01455v1 Announce Type: cross Abstract: Benign fine-tuning severely weakens the safety alignment of large language models (LLMs), so we study why refusal behavior is so fragile. While prior

When the Algorithm Becomes the Brand Crisis: A Sociotechnical Theory of Distributed Responsibility and Accountable Transparency

arXiv:2609.00510v1 Announce Type: new Abstract: Artificial intelligence systems increasingly enact market-facing promises through chatbots, recommendation systems, automated decisions, and generative

When Tokenization is Secretly Output Supervision

arXiv:2609.01386v1 Announce Type: new Abstract: Tokenization in language models is treated by default as an input preprocessing decision. We argue that this framing is incomplete: in autoregressive mo

researcharxiv-cs-cl

Where Should Experience Live? Hierarchical Hebbian Memory for Continual Vision Transformers

arXiv:2609.00358v1 Announce Type: new Abstract: Vision Transformers provide strong visual representations but typically rely on slowly updated parameters, limiting their ability to organize newly acqu

researcharxiv-cs-cv

Where the Verifier Fails: A Category-Level Audit of Reward Signals in RLVR

model-releasesarxiv-cs-cl

arXiv:2609.01354v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) and standard benchmark evaluation both rely on an automatic verifier that turns a free text answer

Who Judges the Judges? A Chinese Safety QA Benchmark for Evaluating LLM Responses and Safety Judges

arXiv:2609.01210v1 Announce Type: cross Abstract: Safety benchmarks for large language models often assess the risk of a user query, although the outcome of question answering depends on whether the r