Search: “arxiv-cs-ai”

Task-Relevant Null-Space Residuals for Non-Injective Neural Mappings

arXiv:2609.37272v1 Announce Type: new Abstract: Non-injective mappings in neural networks map distinct inputs to the same representation, thereby implicitly inducing equivalence relations in the input

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Teaching LLMs to Generate Challenging MILP Instances via Solver Feedback

DGX agent

arXiv:2609.37356v1 Announce Type: new Abstract: Generating optimization instances that are both feasible and computationally challenging is crucial for benchmarking solvers and training learning-based

model-releasesarxiv-cs-ai

30 Sept 2026

Text2Sim: Agentic Physics-Based Simulation Generation with Distilled Expertise

arXiv:2609.36593v1 Announce Type: cross Abstract: Creating diverse physical simulations remains labor-intensive because assets, layout, physical parameters, motion, control, and rendering must be desi

Local Ai

The Commit-Abstain Circuit: Why Language Models Hallucinate Instead of Abstaining

arXiv:2609.32964v2 Announce Type: replace Abstract: Language models (LMs) often hallucinate by committing to confident answers rather than abstaining, even when they do not have enough information to

local-aiarxiv-cs-ai

The Default Trap: Rethinking Plan Evaluation in Tool-Using LLM Agents

arXiv:2609.36829v1 Announce Type: new Abstract: An executor can respond strongly to a change in a supplied plan's priority while showing a small change in the same information-selection probability wh

The Detectability Gap: Hidden Heterogeneity in Hallucination Detection Across Language Models

arXiv:2609.35860v1 Announce Type: cross Abstract: Sampling based consistency is widely used for hallucination detection, yet aggregate performance can conceal systematic differences in which errors ar

The Layer Mystery of VLA: An Information-Theoretical Analysis of VLA Latent Interface

arXiv:2609.36118v1 Announce Type: new Abstract: Vision-language-action (VLA) policies connect a pretrained vision-language backbone to an action head through a latent interface, but which backbone lay

The Price of Token Boundaries: Compression Certificates and Prediction

arXiv:2609.35869v1 Announce Type: new Abstract: Pre-tokenisation restricts which text fragments can become prediction units, but its compression cost is obscured when tokenisers are compared only unde

The Safety Operator: Modulating the Expression of Safety Instructions via Spectral Optimization

arXiv:2609.36434v1 Announce Type: new Abstract: Context tokens in a transformer-based language model can be absorbed into the model's weights as a multiplicative operator. We study this operator in th

The Unequal Influence of Bad Advice: Using Training Data Attribution to Modulate Emergent Misalignment

arXiv:2609.37914v1 Announce Type: cross Abstract: Fine-tuning large language models on narrow, misaligned tasks can undo their post-training alignment and induce novel misaligned behaviors -- a phenom

The Uneven Decline of Collective Knowledge Production: Evidence from Stack Overflow After Generative AI

arXiv:2609.36069v1 Announce Type: cross Abstract: Generative AI (Gen AI) is reshaping how individuals learn and work, but its consequences for collective knowledge, the shared body of knowledge that o

Think Before You Restore: Risk-Aware Manchu Manuscript Restoration with Stroke-Guided Attention

arXiv:2609.36243v1 Announce Type: cross Abstract: Full-page blind restoration of historical Manchu manuscripts is challenging due to scarce annotations, unknown degradation regions, and fragile connec

Thinking Before Thinking: Scaling Agentic Inference Through Meta-Reasoning

arXiv:2609.38147v1 Announce Type: new Abstract: As agents take on longer and more complex problems, controlling the execution becomes a task in its own right. Each step in the run brings new control c

Thinking in Depth, Speaking Directly: Recurrent Latent Reasoning for Paralinguistically Grounded Spoken Dialogue

arXiv:2609.37818v1 Announce Type: cross Abstract: Empathetic spoken dialogue requires models to use both what is said and how it is said to decide how to respond. Explicit CoT can improve paralinguist

ThinkingGuard: Decoding Implicit Hazards via Step-by-Step Risk Attribution in Multimodal Large Language Models

arXiv:2609.36562v1 Announce Type: cross Abstract: While Multimodal Large Language Models (MLLMs) are increasingly deployed in safety-critical domains, their reliability is threatened by multimodal imp

ThinQuant: Scalable Rotation Learning for Weight and Activation Quantization of LLMs

arXiv:2609.36120v1 Announce Type: cross Abstract: Learned rotations play an important role in enabling low-bit weight and activation quantization of large language models by smoothing outliers in the

ThuRunel: Dynamic Decoupling for Structured Advisory Dialogue

arXiv:2609.36340v1 Announce Type: new Abstract: High-stakes advisory domains such as medical aesthetics, legal consultation, and educational planning exhibit a two-phase structure. The early phase req

ToolFence: Fine-Grained Authorization for Secure Tool-Using LLM Agents

arXiv:2609.37196v1 Announce Type: cross Abstract: Tool-using LLM agents remain vulnerable to indirect prompt injection because trusted instructions and untrusted observations share one context, allowi

Topological Coherence for Self-evolving Multi-agent Systems

arXiv:2609.37953v1 Announce Type: new Abstract: Complex tasks inherently couple workflow structure, agent responsibility, collaboration, and memory access: task regions delimit responsibility and tool

TORQUE: Optimizing What (not) to Quantize Before and After Rotation

arXiv:2609.36032v1 Announce Type: cross Abstract: Uniform random rotations are an effective preprocessing step for quantization: they make normalized coordinate distributions approximately Gaussian, e

Towards an AI Software Factory for Data Systems

arXiv:2609.36323v1 Announce Type: new Abstract: AI-assisted coding tools deliver significant acceleration of coding, but only limited impact across the end-to-end software development lifecycle (SDLC)

Towards Breaking the Learning System Wall Using Multimodal Tutoring Transcriptions

arXiv:2609.36502v1 Announce Type: cross Abstract: Past research using log data has faced the 'learning system wall,' whereby few methods exist for generalizing models of student learning across platfo

Towards Mitigating Deceptive Safety Alignment in Large Reasoning Models

arXiv:2609.36254v1 Announce Type: new Abstract: Large Reasoning Models (LRMs) are commonly trained with reinforcement learning (RL) to improve their generation of chain-of-thought (CoT) reasoning befo

TRACE: Deployable Tree-Relational Structure Enhancement for Oncology LLMs

arXiv:2609.35810v1 Announce Type: cross Abstract: Large language models are increasingly used in oncology applications, but their predictions are often weakly grounded in explicit medical structure. W

Train Ahead, Distill Back: Bootstrapping On-Policy Self-Distillation for Large Language Models

arXiv:2609.37132v1 Announce Type: new Abstract: On-policy self-distillation (OPSD) improves large language models by letting a self-teacher with privileged information provide dense token-level superv

Training-Free Global Geometric Association for 4D LiDAR Panoptic Segmentation

arXiv:2512.18991v3 Announce Type: replace-cross Abstract: Dominant paradigms for 4D LiDAR panoptic segmentation are usually required to train deep neural networks with large superimposed point clouds

Training LLMs to Verbalize Evaluation Awareness

arXiv:2609.36316v1 Announce Type: cross Abstract: Evaluation awareness (EA) can cause large language models (LLMs) to behave differently during audits than in deployment, yet measuring and accounting

Trajectory Soup: Pushing the Compute-Scaling Frontier of LLM Mid-training via Diverse Trajectories

arXiv:2609.37169v1 Announce Type: cross Abstract: Mid-training equips pretrained large language models with specialized and reasoning capabilities, but the returns of this stage are bounded since addi

Transformers Stop Thinking Too Early, and a Tiny LoRA Fixes It

arXiv:2609.36585v1 Announce Type: new Abstract: Pretrained transformers use little of their depth to follow references in context. Thirteen base models reliably follow only 1.4-3.6 lines, and extra pr

Transolver-sigma: Joint Spectral-Physical Subspace Modeling for Neural PDE Solving

arXiv:2609.37279v1 Announce Type: new Abstract: Neural solvers offer efficient surrogates for numerical simulation of partial differential equations (PDEs). For time-dependent problems, strong one-ste

TReVS: Integrating Textual Relevance and Visual Saliency for Efficient Vision-Language Model Token Pruning

arXiv:2609.37581v1 Announce Type: cross Abstract: Vision-Language Models (VLMs) excel at visual understanding and reasoning but often incur substantial inference costs due to the large number of visua

UNBIND: UNlearning By INference-time Directional Steering for Code LLMs

arXiv:2609.35913v1 Announce Type: cross Abstract: Code large language models acquire programming capabilities from large code corpora, but can also memorize implementations that later require removal.

Understanding Decision-Making Mechanisms in Neural Routing Solvers

arXiv:2609.36063v1 Announce Type: cross Abstract: Neural Combinatorial Optimization (NCO) has achieved strong empirical success, yet the internal mechanisms driving model decisions remain largely unex

Local Ai

UniAfford: Token-Routed Multitask Learning for Generalizable 2D-3D Affordance Perception

arXiv:2609.37264v1 Announce Type: cross Abstract: Affordance perception aims to localize actionable regions supporting embodied interaction, yet 2D and 3D affordance grounding have evolved as separate

local-aiarxiv-cs-ai

Unlocking the Critic: Reward-Free Policy Optimization for LLM Post-Training

arXiv:2609.37119v1 Announce Type: cross Abstract: Recent approaches to reinforcement learning (RL) post-training for large language models increasingly remove the critic to reduce training instability

UpliftMem: Learning Set-Level Uplift for Agent Memory Retrieval

arXiv:2609.36805v1 Announce Type: new Abstract: Large language model (LLM) agents reuse external memory to guide new tasks, but effective retrieval requires learning which memory sets improve executio

UserProxyBench: Evaluating LLM User Simulators for Agent Benchmarks and Training

arXiv:2609.38043v1 Announce Type: new Abstract: Interactive agent benchmarks and multi-turn reinforcement learning increasingly place a second language model in the role of the user. This simulated us

V-Engram: Trigger-Indexed External Memory for Modular Text-to-Image Personalization

arXiv:2609.37198v1 Announce Type: new Abstract: Pretrained text-to-image models contain broad visual knowledge, yet they cannot reliably acquire or refine a specific visual identity from only a few re

V-JEPA Policy: Building Effective World-Action Models on Predictive Visual Latents

arXiv:2609.37250v1 Announce Type: cross Abstract: World-action models (WAMs) couple future visual-state prediction with action generation. By adapting video generators or image-editing models pretrain

VAA-CSEC: Vote-guided Advantage Allocation for Chinese Semantic Error Correction

arXiv:2609.36804v1 Announce Type: cross Abstract: Chinese Semantic Error Correction (CSEC) targets semantic errors in Chinese text, which are typically more subtle and complex than spelling and gramma

VACE: Validation-Gated Alternating Co-Evolution of Agent Models and Harnesses

arXiv:2609.37105v1 Announce Type: cross Abstract: Language model agents can be improved by updating their model weights or refining the harness that guides task execution. These components are coupled

VeriWeave Govern: Evidence-Gated Deterministic Runtime Governance for Enterprise AI Agents

arXiv:2609.37457v1 Announce Type: new Abstract: Enterprise artificial-intelligence agents increasingly call tools, modify infrastructure, and process protected data, creating a need to separate action

Video-RSI: Recursive Self-Improvement of Video Understanding Agents via Harness Evolution

arXiv:2609.37950v1 Announce Type: new Abstract: Video understanding agents acquire evidence through an executable harness that controls what they observe and how they use those observations. However,

Video2STL: Grounding VLM-Generated Temporal Specifications for Robot Learning

arXiv:2609.37519v1 Announce Type: cross Abstract: Video-based policy learning is particularly promising, as it illustrates target behaviors without requiring action annotations or embodiment-matched d

VIF-Bench: Evaluating Visual Instruction Following in Multi-Reference Image Generation

arXiv:2609.37709v1 Announce Type: cross Abstract: Recent multimodal image generation models can take multiple images and textual instructions as input, enabling reference-based generation guided not o

VISTA-Bench: Benchmarking Multilingual Image Translation with Image-Specific Rubrics

arXiv:2609.37287v1 Announce Type: cross Abstract: Image translation is a fundamental capability of multimodal models for multilingual applications, requiring visual understanding and meaning preservat

VISTA: Value-Informed Event Appraisal for Multimodal Emotion Conflict

arXiv:2609.37324v1 Announce Type: new Abstract: Conflicting emotional cues can be individually valid: a subdued voice may reflect a blocked goal while a smile satisfies a social obligation. Their inte

Visual Parallel Search: Learning to Search High-Resolution Images with Parallel Tile Inspection and Adaptive Zoom

arXiv:2609.37002v1 Announce Type: cross Abstract: High-resolution visual question answering often fails because a multimodal model does not acquire the small, spatially localized evidence needed to an