Search: “llm”

Not Too Hard, Not Too Easy: Learning from Intermediate States for LLM Structured Reasoning

arXiv:2609.33149v2 Announce Type: replace Abstract: A common principle of effective learning is to practice material that is neither already mastered nor too difficult to permit progress. We ask how t

2 Oct 2026

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Model Releases

OpenMTB-Audit: Exposing Over-Refusal and Clinical Expert Perspectives in LLM-Based Molecular Tumor Board Safety Evaluation

DGX agent

arXiv:2610.01497v1 Announce Type: new Abstract: Molecular tumor boards integrate genomic findings, clinical context, and therapeutic evidence to support precision oncology. As AI enters this workflow,

model-releasesarxiv-cs-ai

2 Oct 2026

Beyond Text: LLM-Based Dimensional Emotion Evaluation in Multimodal Dialogue

arXiv:2609.39072v1 Announce Type: cross Abstract: Emotion recognition in conversation has been widely studied, but applying Large Language Models (LLMs) to continuous dimensional emotion evaluation in

CamAgent: An LLM-Agent Framework for Multi-Species Camera-Trap Workflows

arXiv:2609.39112v1 Announce Type: new Abstract: Camera traps accumulated vast, multidimensional data for wildlife monitoring, yet translating raw media archives into meaningful ecological insights rem

agentsarxiv-cs-cv

Diversity Combining for Multi-Path LLM Reasoning

arXiv:2609.38829v1 Announce Type: new Abstract: Multi-path reasoning methods such as self-consistency (SC) sample K reasoning paths and choose the most frequent answer. However, their gains quickly pl

researcharxiv-cs-ai

Hardware

EDGC: Entropy-driven Dynamic Gradient Compression for Efficient LLM Training

arXiv:2511.10333v2 Announce Type: replace-cross Abstract: Training large language models (LLMs) at scale incurs substantial communication overhead, while static gradient compression cannot adapt to gr

hardwarearxiv-cs-ai

Fairness Beyond a Single Run: Training-Seed Variability in Speech LLM Adaptation

arXiv:2609.38976v1 Announce Type: new Abstract: Demographic fairness gaps in automatic speech recognition are almost always reported from a single training run. We fine-tune the Q-former projector and

safetyarxiv-cs-cl

Multi-LLM Collaborative Alignment via Stackelberg Games

arXiv:2609.39076v1 Announce Type: new Abstract: A pool of language models can collaborate and improve collectively by learning from one another's responses. These interactions depend on the instructio

Preemptive LLM Unlearning against Forbidden Capability Acquisition via Gradient Sealing

arXiv:2609.39866v1 Announce Type: new Abstract: Open-weight LLMs are released not only as fixed products but also as substrates for downstream fine-tuning. This openness, however, creates legal and et

TACTIC: Temporal and Context-Aware LLM Tactical Planning for Roadside LiDAR Attacks

arXiv:2609.39969v1 Announce Type: cross Abstract: Physical LiDAR attacks are often evaluated using fixed primitives and manually selected parameters, despite their strong dependence on surrounding tra

local-aiarxiv-cs-ai

AutoLoCo: Communication Efficient Distributed LLM Training via Adaptive Synchronization

arXiv:2609.36662v1 Announce Type: cross Abstract: The pre-training of Large Language Models (LLMs) is increasingly conducted across multiple data centers. As training scales to a larger number of acce

local-aiarxiv-cs-ai

Beyond Semantic Narrowing: Robust and Efficient LLM Watermarking with Hamming Neighborhoods

arXiv:2609.37218v1 Announce Type: cross Abstract: Semantic watermarking improves robustness against watermark removal attacks by embedding detectable signals into sentence-level representations. Howev

researcharxiv-cs-ai

Do LLM Agents Execute the Plans They Declare? From Planning-Mode Declaration to Pattern-Specific Execution

arXiv:2609.38108v1 Announce Type: new Abstract: Large language models (LLMs) enable agents to solve long-horizon tasks by generating a plan and then executing it in an environment. However, successful

Dynamic Optimizations of LLM Ensembles with Two-Stage Reinforcement Learning Agents

arXiv:2502.04492v3 Announce Type: replace Abstract: The advancement of LLMs and their accessibility have triggered renewed interest in multi-agent reinforcement learning as robust and adaptive framewo

safetyarxiv-cs-cl

Evolving Towards Better Codes: LLM-Guided Search for High-Distance Binary Linear Codes

arXiv:2609.37056v1 Announce Type: cross Abstract: Evolutionary program search driven by large language models (LLMs) has produced record-breaking constructions for open problems in combinatorics and b

researcharxiv-cs-ai

Reconstructing Implicit Scientific Knowledge: Evaluating LLM Agents through End-to-End Reproduction of Astronomy

arXiv:2609.35900v1 Announce Type: cross Abstract: The integration of large language models (LLMs) into scientific workflows is accelerating, yet their ability to reconstruct the reasoning underlying p

agentsarxiv-cs-ai

Reliable but Design-Sensitive: Instrument Uncertainty in LLM Annotation

arXiv:2609.35824v1 Announce Type: new Abstract: Large language models (LLMs) can give reliable labels under one setup yet change those labels when researchers make other reasonable design choices. We

researcharxiv-cs-cl

Applications

SemOPT: Fixing Semantic Errors in LLM-based Optimization Modeling via Reward-Guided Search

arXiv:2609.37361v1 Announce Type: new Abstract: Operations research supports decision-making in domains such as energy, economics, and healthcare. Solving operations research problems typically begins

applicationsarxiv-cs-cl

SINGED: Correct Outputs Do Not Certify Safe Execution in LLM Agents

arXiv:2609.35889v1 Announce Type: cross Abstract: Tool-using language-model agents select and execute third-party artifacts. Different implementations can return the requested output while producing h

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

A Cheap Verifier is Good Enough: LLM Post-training is Robust to Erroneous Rewards

arXiv:2609.33467v1 Announce Type: new Abstract: When post-training large language models on tasks with semi-verifiable rewards, there are many factors (training steps, base model size, training order,

AlphaPareto: Formulaic Alpha Discovery with LLM-Guided Multi-Objective Reinforcement Learning

arXiv:2609.34188v1 Announce Type: cross Abstract: Formulaic alpha discovery is a core challenge in quantitative trading, as identifying alphas that work well together remains difficult. Recent reinfor

agentsarxiv-cs-lg

DRIFT: Data Selection for LLM Instruction Tuning via On-Policy Attribution

arXiv:2606.18307v2 Announce Type: replace Abstract: Data selection for instruction tuning is usually studied for efficiency, reaching strong performance with a fraction of the training data. We ask wh

safetyarxiv-cs-lg

From One-Shot Generation to Incremental Music Composition: Adapting a General-Purpose Instruction LLM for Persistent Symbolic Editing

arXiv:2609.34994v1 Announce Type: cross Abstract: Most music-generation systems are still framed and evaluated primarily as producers of complete outputs, whereas composition often proceeds through su

GraphHCA: Closed-Form Hindsight Credit Assignment for Long-Horizon LLM Agents

arXiv:2609.35084v1 Announce Type: new Abstract: Group-based reinforcement learning (RL) has advanced large language models (LLMs) and is increasingly extending to agentic tasks, where sparse terminal

safetyarxiv-cs-lg

PrefillShare: A Shared Prefill Module for KV Reuse in Multi-LLM Disaggregated Serving

arXiv:2602.12029v2 Announce Type: replace Abstract: Multi-agent systems increasingly orchestrate multiple specialized language models to solve complex real-world problems, often invoking them over a s

agentsarxiv-cs-lg

Prioritizing Repeated LLM Evaluation for Hidden Failure Discovery

arXiv:2609.32547v1 Announce Type: new Abstract: Large language models are commonly evaluated by generating a small number of stochastic responses for each prompt in a benchmark. Because inference budg

Understanding Quantization of Optimizer States in LLM Pre-training: Dynamics of State Staleness and Effectiveness of State Resets

arXiv:2603.16731v2 Announce Type: replace Abstract: Quantizing optimizer states is becoming an important ingredient of memory-efficient large-scale pre-training, but the resulting optimizer dynamics r

researcharxiv-cs-lg

WSM-Aware HRI: An IoT-Enhanced Framework for Early Detection and Norm-Guided Repair of Failures with LLM Guidance

arXiv:2609.32336v1 Announce Type: new Abstract: Human-robot interaction (HRI) failures remain a major barrier to deploying robots in real-world environments. Prior work often treats failures as isolat

safetyarxiv-cs-ro

Can someone explain how a LLM goes rogue.

Alright so I'm pretty new to all of this, but today I got a pretty solid crash course in exactly how these LLMs work. I wad quite surprised to find out that there's zero 'thinking' involved in any of

agentsr-chatgpt

28 Sept 2026

Evaluation is All You Need: Strategic Overclaiming of LLM Reasoning Capabilities Through Evaluation Design

arXiv:2506.04734v3 Announce Type: replace Abstract: Reasoning models represented by the Deepseek-R1-Distill series have been widely adopted by the open-source community due to their strong performance

28 Sept 2026

What Do They Fix? LLM-Aided Categorization of Security Patches for Critical Memory Bugs

arXiv:2509.22796v2 Announce Type: replace-cross Abstract: Open-source software projects are foundational to modern software ecosystems, with the Linux kernel standing out as a critical exemplar due to

researcharxiv-cs-lg

28 Sept 2026

Beyond Average Safety: Chance-Constrained LLM Fine-tuning

arXiv:2609.29960v1 Announce Type: cross Abstract: Fine-tuning large language models on new objectives can improve helpfulness, instruction following, or domain-specific performance, but it can also in

Design and Evaluation of LLM Chaining-Based Task Planning for General Purpose Service Robots

arXiv:2609.29043v1 Announce Type: cross Abstract: General Purpose Service Robot (GPSR) tasks, as defined in the RoboCup@Home benchmark, require robots to interpret diverse natural language commands an

Epistemic-Probabilistic Model for Guarded Multi-Agent LLM Coordination

arXiv:2609.29366v1 Announce Type: new Abstract: Multi-agent large language models (LLMs) have become ubiquitous in applied AI, yet their theoretical foundations remain surprisingly understudied. Viewe

agentsarxiv-cs-ai

LLM surprisal is necessary but not sufficient to capture English garden-path effects: Evidence from joint latent modeling of reading paradigms

arXiv:2602.04489v2 Announce Type: replace Abstract: Temporarily ambiguous garden-path sentences ('While the team trained the striker wondered... ') are known to cause processing difficulty, which can

researcharxiv-cs-cl

NumericJev: Jev-like LLM Numerical Decoding with Multiway Decision Trees

arXiv:2609.28587v1 Announce Type: cross Abstract: Large language models can interpret natural lan- guage, yet robust decisions remain challenging. Jev-like models expose structured choices, but these

Polite but Misaligned: Evaluating LLM Politeness Judgments Against Human Pragmatic Norms

arXiv:2609.29001v1 Announce Type: new Abstract: Despite strong performance on standard benchmarks, it remains unclear whether large language models (LLMs) evaluate social pragmatics in ways that align

safetyarxiv-cs-cl

PrivDrift: Auditing User-Secret Leakage Under Topic Drift in Active LLM Conversations

arXiv:2609.30094v1 Announce Type: new Abstract: Large language models increasingly operate as persistent assistants in user-facing, shared-session, and tool-augmented settings. When users disclose sen

Stale Does Not Mean Unsafe: Guard Precision for Tool-Using LLM Agents under Infrastructure State Races

arXiv:2609.29522v1 Announce Type: new Abstract: Tool-using language-model agents increasingly mutate schedulers, data pipelines, object stores, and access-control systems. Between an agent's read and

Agentic Governance and Adversarial Verification for Policy-Constrained LLM Healthcare Appeal Generation

arXiv:2609.27844v1 Announce Type: new Abstract: Claim denial management costs U.S. healthcare approximately $260 billion annually in administrative overhead. Large Language Models (LLMs) and Retrieval

local-aiarxiv-cs-ai

Crossflow: Prefill-Decode Elasticity for Agentic LLM Serving

arXiv:2609.27085v1 Announce Type: cross Abstract: As serving capacity demand surpasses that of training, serving efficiency becomes increasingly important. Prefill-decode (P/D) disaggregation improves

agentsarxiv-cs-ai

FDE-Bench: Evaluating LLM Agents for Deployment Environment Configuration

arXiv:2609.27571v1 Announce Type: cross Abstract: Deployment requires an agent to turn application code into a running system whose services connect, become ready, and remain observable. FDE-Bench eva

MacBook Pro M5 Max LSE LLM running an AMD Radeon AI PRO R9700 over Thunderbolt 5 in a Razer enclosure

Hello Everyone! Good news for anyone who can compile code on apple, you can now experience using a AMD Radeon AI Pro R9700 on a MacBook Pro or Mac mini, Mac Studio or any OSX variant as long as it has

local-air-localllama

Policy-as-Skill: Governed LLM Decision Support with Evidence, Deterministic Control, and Audit

arXiv:2609.27087v1 Announce Type: new Abstract: Organizations increasingly use LLMs for policy, compliance, risk, and operational decision support, requiring evidence validation, review routing, versi

Safe Multi-Robot Coordination via VLM-LLM Reasoning and Reachability Analysis

arXiv:2609.27816v1 Announce Type: cross Abstract: Safe coordination in heterogeneous machine-to-machine (M2M) robotic systems is challenging when robots differ in sensing capabilities, environmental a

SR-Fraud: An Outcome-Supervised Reflective LLM Agent Framework for Non-Stationary Payment Fraud Detection

arXiv:2609.27287v1 Announce Type: new Abstract: Real-time payment fraud detection is a non-stationary streaming prediction problem: adversaries adapt before supervised labels mature, and localized bur

APEXA: Execution-Integrity Enforcement for Multi-Agent LLM Automation of Synchrotron Data Reduction

arXiv:2609.24165v1 Announce Type: new Abstract: Synchrotron data reduction, detector calibration followed by azimuthal integration of terabyte-scale diffraction series, is a multi-step, expert-bound b