Search: “llm”

BRIDGE: Bilevel Retrieval-Credit-Aware Agentic Reinforcement Learning

arXiv:2609.36505v1 Announce Type: new Abstract: Agentic reinforcement learning (ARL) with verifiable rewards improves the ability of large language models (LLMs) to tackle knowledge-intensive tasks by

EasyPPO: Stabilizing the Critic Is Key

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

DGX agent

arXiv:2609.36802v1 Announce Type: new Abstract: A key strength of Proximal Policy Optimization (PPO) is its learned critic, which uses historical trajectories collected during reinforcement learning t

safetyarxiv-cs-lg

30 Sept 2026

Environment Steering: Using Data Flow Control to Improve Agent Utility and Safety

arXiv:2609.35807v1 Announce Type: cross Abstract: LLM agents can make unsafe tool calls even when instructed to behave safely. Existing defenses constrain agents before execution, modify tool inputs/o

GeoOutageBench: Benchmarking Ambiguity-aware, Ontology-grounded Geospatiotemporal KGQA for Multimodal Power Outage and Resilience Analysis

arXiv:2609.36082v1 Announce Type: new Abstract: We introduce GeoOutageBench, a benchmark for assessing LLM-based geospatiotemporal KGQA for multimodal outage and resilience analysis. Unlike existing K

HEAR: Real Voices, Real Bias: A Large-Scale Human-Recorded, Demographically Diverse Benchmark for Audio Language Models

arXiv:2609.35952v1 Announce Type: cross Abstract: We introduce HEAR (Human-recorded Evaluation of Audio-LLM bias by Real speakers), a large-scale, ecologically valid benchmark comprising 87k real huma

KV-Kaizen: Learning Context-Adaptive Cache Compression Choices

arXiv:2609.37988v1 Announce Type: new Abstract: As the context size of text processed with an LLM grows, the size of KV caches can outstrip the memory allocated for the original model weights. This im

local-aiarxiv-cs-ai

Lost in Conversation or Lost in Translation? Diagnosing Multi-Turn Degradation in RAG

arXiv:2609.36700v1 Announce Type: cross Abstract: When conversing with large language models (LLMs), users often begin with a simple question and build towards a multi-hop question through follow-up t

On Calibration of Large Language Models: From Response To Capability

arXiv:2602.13540v2 Announce Type: replace-cross Abstract: Accurate confidence estimation is critical for reliable use of large language models (LLMs). Prior work on LLM calibration largely focuses on

OTROPE: Optimal Transport-based Robust Off-policy Evaluation for Large Language Models

arXiv:2609.36264v1 Announce Type: new Abstract: Reliable evaluation of large language models (LLMs) is essential for their development and deployment, yet is often costly, risky, and difficult to perf

AgentWare: Automating the Lifecycle of Agentic Applications across the Edge-to-Cloud Continuum

arXiv:2609.34586v1 Announce Type: cross Abstract: Deploying LLM-enabled agentic applications across the Edge-to-Cloud continuum remains challenging due to hardware heterogeneity, deployment complexity

agentsarxiv-cs-lg

Cartridges++: KV Cache Compression without Off-Context Derailment

arXiv:2609.35621v1 Announce Type: new Abstract: Serving long documents to a Large Language Model (LLM) repeatedly is expensive: computations grow with context length, and the memory footprint of the k

researcharxiv-cs-lg

NetInjectBench: Benchmarking Indirect Prompt Injection in Tool-Using Large Language Model Agents for Network Operations

arXiv:2607.10490v2 Announce Type: replace-cross Abstract: Tool-using large language model (LLM) agents are attractive for network operations, but tickets, alerts, logs, runbooks, and ChatOps messages

P^2O: Joint Policy and Prompt Optimization

arXiv:2603.21877v4 Announce Type: replace Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) enhances Large Language Model (LLM) reasoning but is suffer from advantage collapse: when all

Self-Evolving Time-Series Forecasting Agents with Episodic Memory and Online Policy Learning

arXiv:2609.32689v1 Announce Type: new Abstract: LLM-based agents are increasingly used for time-series forecasting because they can organise contextual information, perform multi-step analysis, and gu

Silent Failures in Agentic Security Evaluation: A Validated Harness for Tool-Call Mediation Under Indirect Prompt Injection

arXiv:2609.32691v1 Announce Type: cross Abstract: LLM agents that invoke privileged tools are vulnerable to indirect prompt injection (IPI), in which adversarial instructions embedded in retrieved dat

When Keywords Drop but Classifiers Hold: Soft Refusals under KV Cache Compression

arXiv:2609.31678v1 Announce Type: new Abstract: KV cache compression is widely used for long context LLM inference under memory constraints, while deployed systems typically score refusals after gener

Acoustic-to-Text KV Compression for Full-Duplex Speech Models

arXiv:2609.31224v1 Announce Type: cross Abstract: Full-duplex speech language models continuously accumulate acoustic key-value (KV) states, making long-running interactions memory-intensive. During l

Tutorials

Aurora-X: Built for Extreme Time Series Forecasting

arXiv:2609.31038v1 Announce Type: new Abstract: Time series foundation models (TSFMs) enable cross-domain forecasting, but their development as general-purpose forecasters remains constrained by under

tutorialsarxiv-cs-lg

CARGO: Context-Aware Retrieval-Gated Evaluation of Agentic AI in Production

arXiv:2609.30471v1 Announce Type: cross Abstract: Reference-based LLM-as-a-judge evaluation assumes the reference answer is the target. In deployed agentic systems that operate over dynamic entities (

agentsarxiv-cs-ai

ConsultMind:Towards Automated Diagnostic Consultation via Uncertainty-Aware Reasoning

arXiv:2609.30796v1 Announce Type: new Abstract: Diagnostic consultation is an online sequential decision-making process in which clinicians gather evidence through patient interaction until a diagnosi

Cosine Similarity Is Not Evidence: Measuring the Noise Floor of Interpretability Transfer Under Quantization

arXiv:2609.30275v1 Announce Type: new Abstract: A statistic reported without the quantity needed to interpret it is not evidence. We develop that thesis for a concrete practice in AI safety. Interpret

Mutable Transcripts: Mitigating Context Pollution through Editable Conversation State

arXiv:2609.31354v1 Announce Type: new Abstract: Contemporary large language model (LLM) chat systems treat conversation history as an immutable sequence of turns that defines the model's working conte

RAPTOR: Ridge-Adaptive Logistic Probes

arXiv:2602.00158v3 Announce Type: replace-cross Abstract: Probing studies what information is encoded in a frozen LLM's layer representations by training a lightweight predictor on top of them. Beyond

SLMFix: Leveraging Small Language Models for Domain Specific Language Error Fixing with Reinforcement Learning

arXiv:2511.19422v2 Announce Type: replace-cross Abstract: Large language models (LLMs) have shown impressive capabilities in code generation across many programming languages but even state-of-the-art

SPO: Discovering Adaptive Large Neighborhood Search Operators via Stackelberg Program Optimization

arXiv:2609.31179v1 Announce Type: new Abstract: Large neighborhood search (LNS) relies critically on destroy and repair operators, whose effectiveness depends on both adaptation to the evolving LNS st

Applications

Where Does Retrieval-Based Open-Ended Evaluation Fail? Automatic Taxonomy Induction from Long-Form Medical Answer Factuality Verification

arXiv:2609.30467v1 Announce Type: new Abstract: Retrieval-based factuality evaluation, where LLM-generated claims are verified against evidence from authoritative medical corpora, has become the domin

applicationsarxiv-cs-cl

Don't Read the Log: Execution Traces Contaminate Verifiers in Video-Generation Agents

arXiv:2609.28564v1 Announce Type: cross Abstract: Agentic video-generation systems close a loop between a generator and a verifier: an LLM plans shots, calls a text-to-video model, and a multimodal ju

25 Sept 2026

On the Effectiveness of Kernel-Level Evidence for Agent Security

arXiv:2609.28915v1 Announce Type: cross Abstract: LLM agents are deployed into infrastructure that grants them broad host authority, yet existing agent-security benchmarks and defenses operate almost

agentsarxiv-cs-ai

25 Sept 2026

EnSIMem: Entity-Structured Indexing for Long-Term Agent Memory

arXiv:2609.27279v1 Announce Type: new Abstract: An agent that interacts with users over long periods must recall facts, preferences, events, and changes from a continuously growing interaction history

agentsarxiv-cs-ai

24 Sept 2026

Agent0: Unleashing Self-Evolving Agents from Zero Data via Tool-Integrated Reasoning

arXiv:2511.16043v2 Announce Type: replace Abstract: Large Language Model (LLM) Agents, often trained with Reinforcement Learning (RL), are constrained by a dependency on human-curated data, limiting s

agentsarxiv-cs-lg

23 Sept 2026

STRUCTUREDAGENT: Planning with AND/OR Trees for Long-Horizon Web Tasks

arXiv:2603.05294v3 Announce Type: replace Abstract: Existing LLM-based web agents struggle on complex, long-horizon tasks due to limited in-context memory, weak planning abilities, and greedy behavior

23 Sept 2026

Toward Personalized Sleep Guidance from Wearable Data Using Language Models

arXiv:2609.22463v1 Announce Type: new Abstract: Sleep monitoring using wearable data has shown promise for personal health, yet large language model (LLM)-based summarization and question answering re

local-aiarxiv-cs-cl

22 Sept 2026

Cultural Alignment in Large Language Models Using Soft Prompt Tuning

arXiv:2503.16094v2 Announce Type: replace Abstract: Large Language Model (LLM) alignment is commonly achieved through supervised fine-tuning or reinforcement learning, both of which require labeled or

local-aiarxiv-cs-cl

21 Sept 2026

GraphSkillEvo: Evolutionary Optimization of Graph-Structured Agent Skills

arXiv:2609.21749v1 Announce Type: new Abstract: Skills can improve the performance of Large Language Model (LLM) agents by providing task-specific procedural guidance, while skill optimization further

21 Sept 2026

Predictable Failure in Multi-Hop Retrieval: Score-Distributional Confidence Scoring and Abstention

arXiv:2609.22056v1 Announce Type: cross Abstract: Multi-hop retrieval failures are not uniformly distributed across queries: they cluster in structurally predictable subpopulations. We prove two resul

safetyarxiv-cs-cl

21 Sept 2026

Large Language Models as Falsifiers for Cyber-Physical Systems

arXiv:2609.20752v1 Announce Type: cross Abstract: Falsification searches for counterexamples to formal specifications in cyber-physical systems (CPS). With specifications written in Signal Temporal Lo

MAGS: Multi-agent Auto-formalization Guarantees Safety for Agentic Outputs

arXiv:2609.19391v1 Announce Type: new Abstract: LLM coding agents now generate complex programs at a scale that makes thorough human review increasingly difficult, raising the risk of safety and secur

Message capacity and claim wording set the transition points of collective truth-finding in language-model networks

arXiv:2609.19183v1 Announce Type: cross Abstract: Whether human or large language model (LLM), an agent in a discussion reads only a few of the others' contributions, bounded by cognition, context, or

Q&A on Any Spreadsheet Requires Interpreting Its Grid Structure

arXiv:2609.20732v1 Announce Type: new Abstract: Semantic cell annotation improves chunking interpretability for spreadsheets in LLM-driven RAG systems, aiding answer generation through enriched contex

REARL: A Closed-loop Autonomous Driving Simulation Enhancement Framework with Real Traffic Data and Large Language Models

arXiv:2609.19903v1 Announce Type: new Abstract: Accurate simulation is crucial for autonomous driving development, yet capturing real-world traffic complexity remains challenging. Existing simulators

agentsarxiv-cs-lg

To Copy or Not to Copy: Controlling Speculative Decoding via Intrinsic Model Signals

arXiv:2609.20186v1 Announce Type: new Abstract: Speculative Decoding (SD) has significantly accelerated Large Language Model (LLM) inference, yet existing approaches face a fundamental tradeoff betwee

Understanding and Exploiting Diagonal Attention Sparsity in Autoregressive Image Generation

arXiv:2609.19702v1 Announce Type: new Abstract: Autoregressive image generation has emerged as a paradigm for multimodal AI systems due to its compatibility with transformer-based LLM serving infrastr

local-aiarxiv-cs-cv

Autonomy in Check: Governor-Mediated Adaptive Security at the Edge

arXiv:2609.18338v1 Announce Type: cross Abstract: Adaptive security at the network edge increasingly relies on automated planners, including rule-based controllers, learned policies, and LLM-assisted

Correlation-Guided Encoder Selection for Multi-Encoder Large Audio-Language Models

arXiv:2609.18041v1 Announce Type: cross Abstract: Multi-encoder fusion extends Large Audio-Language Models (LALMs) beyond speech-centric recognition, but selecting encoders via intuition or exhaustive

Designing Agentic AI Workflow Portfolios under Imperfect Selection and Compute Cost

arXiv:2609.18126v1 Announce Type: new Abstract: Agentic AI systems often approach the same task through multiple workflows that differ in reasoning strategy, verification structure, and compute cost.

Applications

Emotion Experience, Expression, and Perception: Emotion Analysis on Multimodal Social Media Posts

arXiv:2609.18385v1 Announce Type: new Abstract: Emotions are an essential aspect of human communication, particularly on social media, where authors frequently combine text and images to convey their

applicationsarxiv-cs-cl

GYROval: A Robust Benchmark for Cultural Value Orientation in Large Language Models

arXiv:2609.18384v1 Announce Type: cross Abstract: We present a robust benchmark for measuring cultural value orientation in large language models on the two Inglehart-Welzel axes over several domains

How AI Assistants Respond to Repeated Abuse

arXiv:2609.17547v1 Announce Type: new Abstract: AI assistants are expected to remain useful during difficult interactions, but little is known about how repeated verbal abuse changes their engagement