All entries

Taming System Complexity: Demystifying Software Engineering Agents in Diagnosing Linux Kernel Faults

arXiv:2505.19489v2 Announce Type: replace Abstract: The Linux kernel is a critical system, serving as the foundation for numerous systems. Bugs in the Linux kernel can cause serious consequences, affe

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Research

Taming the Loss Landscape of PINNs with Noisy Feynman-Kac Supervision: Operator Preconditioning and Non-Asymptotic Error Bounds

DGX agent

arXiv:2606.00643v1 Announce Type: cross Abstract: Physics-Informed Neural Networks (PINNs) often train slowly or fail to converge on challenging partial differential equations (PDEs), a behavior recen

researcharxiv-cs-lg

2 Jun 2026

TAP-JEPA: Frozen Future-Latent Probing and Two-Stage Score Fusion for EPIC-KITCHENS-100 Action Anticipation

arXiv:2606.00662v1 Announce Type: new Abstract: This report presents TAP-JEPA, our runner-up submission to the EPIC-KITCHENS-100 (EK-100) Action Anticipation Challenge at EgoVis 2026. The task is to a

TAPS: Target-Aware Prefix Tree Selection for Diffusion-Drafted Speculative Decoding

arXiv:2606.00487v1 Announce Type: new Abstract: Using a diffusion model for parallel drafting is a promising approach for speculative decoding. By predicting tokens at multiple future positions in a s

Local Ai

Target localization, identification and sensing using latent symmetries

arXiv:2606.01421v1 Announce Type: new Abstract: We show that an array of scatterers which has been designed to have latent ('hidden') symmetries can be used as a sensor. We use the capacitance matrix

local-aiarxiv-cs-lg

Task diversity produces systematic transfer but inhibits continual reinforcement learning

arXiv:2606.00880v1 Announce Type: cross Abstract: Continual reinforcement learning aims to produce agents that learn not only to improve at their current tasks but also to adapt as task distributions

Task-Induced Representational Invariances Depend on Learning Objective in Deep RL

arXiv:2606.01868v1 Announce Type: new Abstract: Reinforcement Learning (RL) has long served as a model for goal-directed animal behavior in neuroscience. Modern deep RL has shown remarkable success ac

safetyarxiv-cs-lg

Task Structure Reverses Layerwise State Encoding in Sequence Models

arXiv:2606.00926v1 Announce Type: cross Abstract: Mechanistic studies of sequence models often treat layerwise state encodings as architectural traits: recurrent models concentrate readable state, att

researcharxiv-cs-cl

TCAR-Gen: Temporal Graph Retrieval with Evidence Fusion for Knowledge-Grounded Generation

arXiv:2606.00029v1 Announce Type: cross Abstract: Retrieval-augmented generation systems struggle with temporal reasoning and evidence fusion when answering complex questions over historical criminal

TECCI: Tricky Edits of Collected and Curated Images

arXiv:2606.01213v1 Announce Type: cross Abstract: Despite tremendous recent progress, current text-guided image editing methods still struggle with many aspects of editing involving instruction follow

Agents

TechGraphRAG: An Agentic Graph-Augmented RAG Framework for Technical Literature Reasoning

arXiv:2606.01613v1 Announce Type: cross Abstract: This paper presents an agentic retrieval-augmented generation (RAG) framework for domain-specific technical reasoning support, instantiated over a cur

agentsarxiv-cs-ai

Tempora: Characterising the Time-Contingent Utility of Online Test-Time Adaptation

arXiv:2602.06136v2 Announce Type: replace-cross Abstract: Test-time adaptation (TTA) offers a compelling remedy for machine learning (ML) models that degrade under domain shifts, improving generalisat

Temporal Evidence Routing with Structured Visual Evidence for TimeLogicQA

arXiv:2606.01106v1 Announce Type: new Abstract: TimeLogicQA evaluates whether video question answering systems can reason over temporal relations such as event existence, ordering, persistence, bounda

Temporal Motif Signatures for Temporal Graph Neural Networks

arXiv:2606.01176v1 Announce Type: new Abstract: Real temporal interaction streams carry predictive structure in short-horizon motif patterns -- repetition, reciprocity, star diversity, triadic flow --

Temporally-Aligned Evaluation for Audio-Driven Talking Head Generation

arXiv:2606.01031v1 Announce Type: cross Abstract: Audio-driven talking-head generation has advanced rapidly, yet existing evaluation protocols mainly rely on frame-wise metrics that assume strict temp

TempRet: Temporal Enhancement and Two-Stage Reranking for CVPR 2026 EPIC-KITCHENS-100 Multi-Instance Retrieval Challenge

model-releasesarxiv-cs-cv

arXiv:2605.24470v2 Announce Type: replace Abstract: Video-text retrieval has witnessed remarkable progress driven by large-scale vision-language pretraining, yet most existing approaches inherit an im

TERRA: Task-Embedded Reasoning and Representation Architecture for Cross-Domain Applications

arXiv:2606.01520v1 Announce Type: new Abstract: A single action-conditioned latent predictive architecture can in principle be trained on the structured state of a driving scene, a robot workspace, or

Test-Time Training for Zero-Resource Dense Retrieval Reranking

arXiv:2606.01070v1 Announce Type: cross Abstract: Dense retrievers excel at first-stage candidate generation but lack effective reranking in zero-resource settings. Existing approaches face a fundamen

Tether-Aware Dynamic Collision Avoidance for USV-HROV Systems

arXiv:2606.01112v1 Announce Type: new Abstract: Heterogeneous marine robotic systems composed of an unmanned surface vehicle (USV) and a hybrid remotely operated vehicle (HROV) have shown great potent

safetyarxiv-cs-ro

TextFake: Benchmarking AI-Generated Image Detection on Text-Rich Images

model-releasesarxiv-cs-cv

arXiv:2606.01050v1 Announce Type: new Abstract: Recent AI-generated image (AIGI) detectors perform well on natural-image benchmarks, but their behavior on text-rich forgeries, such as fabricated scree

The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer

arXiv:2602.02557v2 Announce Type: replace-cross Abstract: Recent advances in end-to-end trained omni-models have substantially improved audio capabilities by strengthening text-audio modality alignmen

safetyarxiv-cs-ai

The Assistant as a Privileged Persona: A canonical reference in cross-persona self-recognition

model-releasesarxiv-cs-lg

arXiv:2606.00545v1 Announce Type: new Abstract: Post-trained language models can recognize their own outputs from a sentence or two out of context. In a companion paper itep{jack2026twomodes} we showe

The Case for Model Science: Verify, Explore, Steer, Refine

arXiv:2606.01189v1 Announce Type: new Abstract: We argue that the AI community is now ready to move beyond benchmarking and consolidate scattered efforts in model analysis into a systematic discipline

Agents

The Deterministic Horizon: When Extended Reasoning Fails and Tool Delegation Becomes Necessary

arXiv:2606.00376v1 Announce Type: new Abstract: Extended chain-of-thought reasoning can degrade performance on deterministic state-tracking tasks, not due to preference biases, but limits rooted in th

agentsarxiv-cs-ai

The Entropic Signature of Class Speciation in Diffusion Models

arXiv:2602.09651v2 Announce Type: replace-cross Abstract: Diffusion models do not recover semantic structure uniformly over time. Instead, samples transition from semantic ambiguity to class commitmen

The Geometry of Grokking: Norm Minimization on the Zero-Loss Manifold

arXiv:2511.01938v3 Announce Type: replace-cross Abstract: Grokking is a puzzling phenomenon in neural networks where full generalization occurs only after a substantial delay following the complete me

The Ghost Couple: Correlated LLM Name Priors and Their Haunting of the Web and Academic Publishing

model-releasesarxiv-cs-lg

arXiv:2606.02184v1 Announce Type: cross Abstract: These names do not exist. Elena Vasquez and Marcus Chen have appeared as volcano experts, astronauts, thriller protagonists, podcast hosts, and academ

The Harsh Truth: Segment-Level Analysis of Harsh Driving Events in Milan Using Large-Scale Telematics, Street Networks, and Google Street View

arXiv:2606.00261v1 Announce Type: new Abstract: Police-reported crash statistics remain the standard input for urban road-safety assessment, but their incompleteness and reporting lag limit their usef

safetyarxiv-cs-cv

The Image Reconstruction Game: Drawing Common Ground Through Iterative Multimodal Dialogue

arXiv:2606.01901v1 Announce Type: cross Abstract: We introduce the Image Reconstruction Game, a fully automated benchmark in which a vision-language model issues corrective instructions to an image ge

The Invisible Coalition Partner: How LLMs Vote When Democracy Gets Concrete

model-releasesarxiv-cs-cl

arXiv:2606.00048v1 Announce Type: cross Abstract: Prior research has established that instruction-tuned large language models exhibit left-of-center political bias, measured exclusively through abstra

The Lie We Tell: Correcting the Euclidean Fallacy in Vision Language Action Policies via Score Matching on Tangent Space

arXiv:2606.01847v1 Announce Type: cross Abstract: Diffusion-based Vision-Language-Action policies achieve remarkable success in robotic manipulation, yet commit a fundamental geometric error we term t

The New Social Image: How AI Competency and AI Proactivity Influence Self- and Peer-Perceptions in the Workplace

arXiv:2606.00182v1 Announce Type: cross Abstract: Human-AI collaboration is considered the most promising way to incorporate AI in the workplace. What remains unexplored are the experiential consequen

The Paradox of Outcome Optimization: A Causal Information-Theoretic Bound on Reasoning Shortcuts in LLMs

arXiv:2606.00674v1 Announce Type: cross Abstract: Large Language Models (LLMs) aligned via outcome-based Reinforcement Learning (RL) frequently exhibit a critical failure mode: they achieve high perfo

safetyarxiv-cs-ai

The Refusal--Compliance Tradeoff: A Large-Scale Safety Behavior Audit of Large Language Models

arXiv:2605.05427v2 Announce Type: replace Abstract: Refusal rates are a poor proxy for LLM safety, i.e., a model may over-refuse benign prompts while still complying with harmful ones. We audit both f

The Representation-Rationalizability Tradeoff in Reward Learning

arXiv:2606.00291v1 Announce Type: cross Abstract: In RLHF, each training example contains a prompt x and two candidate responses y,y', and annotators provide pairwise preferences between these respons

The Right Inference Strategy Is All You Need: Nearly Training-Free Domain-Wise Inference for EgoCross Challenge

arXiv:2606.00829v1 Announce Type: new Abstract: EgoCross evaluates multimodal large language models on egocentric video question answering under substantial domain shift, where test videos come from s

The Role of Ambiguity in Error Prediction via Uncertainty Quantification

arXiv:2606.02093v1 Announce Type: cross Abstract: The task of Error Prediction, namely predicting whether a model output is correct, is commonly tackled with Uncertainty Quantification (UQ). However,

The role of class encoding in neural collapse

arXiv:2606.00344v1 Announce Type: new Abstract: Neural collapse is a structural property of the last-hidden-layer activations in neural network classification models, when trained beyond a zero classi

safetyarxiv-cs-lg

The Shape of Wisdom: Decision Trajectories in Language Models

arXiv:2606.01202v1 Announce Type: new Abstract: Language models do not simply choose an answer at the output layer. In a 9,000-trajectory MMLU study across Qwen2.5-7B-Instruct, Llama-3.1-8B-Instruct,

The Social Cost of Intelligence: Emergence, Propagation, and Amplification of Stereotypical Bias in Multi-Agent Systems

arXiv:2510.10943v2 Announce Type: replace-cross Abstract: Bias in large language models (LLMs) remains a persistent challenge, often leading to stereotyping and unfair treatment across social groups.

safetyarxiv-cs-cl

Theoretical Analysis of Engression and Reverse Markov Engression

arXiv:2606.01002v1 Announce Type: cross Abstract: Engression is a recently proposed and effective framework for conditional distribution learning. Its multi-step Reverse Markov extension further impro

Thinking Economically: A Hierarchical Framework for Adaptive-Complexity Reasoning in LLMs

arXiv:2606.01168v1 Announce Type: new Abstract: Chain-of-Thought (CoT) has significantly enhanced LLM reasoning, yet often incurs substantial computational overhead due to 'overthinking': generating e

researcharxiv-cs-cl

Agents

Thinking in Blender: Staged Executable Inverse Graphics with Vision-Language Models

arXiv:2606.02580v1 Announce Type: new Abstract: Inverse graphics is a longstanding and highly underconstrained problem that seeks to reconstruct images as editable 3D scenes which can be rendered, rel

agentsarxiv-cs-cv

ThinkSwitch: Context Distillation with LoRA and Weight Interpolation for Specific-Purpose Reasoning Tasks

arXiv:2606.01080v1 Announce Type: cross Abstract: Large language models often improve on difficult tasks by spending inference-time compute on a reasoning trace before producing the final answer. That

THRD: A Training-Free Multi-Turn Defense Framework for Jailbreak Attacks on Large Language Models

arXiv:2606.01738v1 Announce Type: cross Abstract: Multi-turn jailbreak attacks pose a growing threat to LLMs by exploiting conversational dynamics such as gradual escalation and cross-turn coordinatio

safetyarxiv-cs-ai

Threading Optimization for Vision-Language-Action Model Inference in Low-Cost Smart Agricultural Manipulation

arXiv:2606.00966v1 Announce Type: new Abstract: Vision-Language Action (VLA) models continue to face challenges such as slow inference speed and difficulty performing fine-grained motion adjustments,

safetyarxiv-cs-ro

Hardware

Threshold-Based Exclusive Batching for LLM Inference

arXiv:2606.00516v1 Announce Type: new Abstract: Mixed batching (MB)--interleaving prefill and decode in a single batch--has become the standard scheduling strategy for large language model (LLM) infer

hardwarearxiv-cs-ai

Through the PRISM: Principle-Aware, Interpretable, and Multi-Scale Evaluation of Visual Designs

model-releasesarxiv-cs-cv

arXiv:2606.00592v1 Announce Type: new Abstract: Effective visual communication stems from the harmony of multiple design principles, such as readability, contrast, alignment, overlap, and coherence, w