Search: “llm”

Are Stated Reasoning Steps Causally Load-Bearing?

arXiv:2609.27038v1 Announce Type: new Abstract: Chain-of-thought (CoT) monitoring assumes that the reasoning a model writes reflects the computation that directly produces its answer. Previous faithfu

researcharxiv-cs-ai

Benchmarking Text-to-SQL under Role-Based Access Control

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

arXiv:2607.22115v2 Announce Type: replace-cross Abstract: Given a database S and a natural language question Q, text-to-SQL systems aim to generate an SQL query that correctly answers Q when executed

model-releasesarxiv-cs-ai

24 Sept 2026

Beyond Poetry: Can Large Language Models Generate Classical Arabic Maqamat?

arXiv:2609.28245v1 Announce Type: cross Abstract: Large language models (LLMs) have shown strong performance in creative text generation, yet their ability to produce culturally grounded and stylistic

ChipMEM: Verification-Grounded Memory for EDA Agents

arXiv:2609.27067v1 Announce Type: cross Abstract: Large language model (LLM)-based agents use Electronic Design Automation (EDA) tools to generate and revise register-transfer-level (RTL) designs unde

Applications

ContraVis: Evidence-Grounded Visual Analytics for Contradiction Review in Legal Contracts

arXiv:2609.27014v1 Announce Type: cross Abstract: Legal contracts are structurally complex documents in which contradictions may emerge across distant and interconnected provisions. Although large lan

applicationsarxiv-cs-cl

Learning What to Activate: Combinatorial Capability Allocation for Long-Horizon Multimodal Agents

arXiv:2609.27869v1 Announce Type: new Abstract: Long-horizon multimodal agents rely on specialized capabilities for perception, retrieval, reasoning, verification, and execution. Existing designs typi

safetyarxiv-cs-ai

Resource-Adaptive Stochastic Gradient Descent for Online Linear Programming without Re-solving

arXiv:2609.28263v1 Announce Type: new Abstract: The growth of large language model (LLM) inference and search services increases the scale of online linear programming problems, motivating computation

researcharxiv-cs-lg

Shutdown Sabotage Propensities in Multi-Agent Systems

arXiv:2609.28274v1 Announce Type: new Abstract: The final safeguard against rogue AI behavior is the human ability to shut systems down. It has been theorized that when an AI is instructed to perform

Local Ai

The 'Alignment Tax' is ruining agentic workflows. Anyone else noticing 'Complexity Theater' in recent models?

Been building some multi-agent pipelines recently and the degradation in direct execution is insane. Whenever I pass a strict operational variable or hypothesis, the LLM triggers this 'Devil's Advocat

local-air-ollama

The Like Trap: Multi-Stage Poisoning against Agents in Similarity-based Recommendation Systems

arXiv:2609.27155v1 Announce Type: cross Abstract: With recent advancements in large language models (LLMs) and LLM-based agents, these agents are becoming increasingly autonomous and gaining broader a

The Tasteful Agent: Measuring and Improving Taste in Long-Horizon Tasks

arXiv:2609.25804v2 Announce Type: replace Abstract: LLM agents increasingly work on long-horizon tasks, and the decisions they make along the way, such as which hypothesis to test or which implementat

TwinCheck: Evidence-Grounded Negative-Twin Verification for Stateful Tool Agents

arXiv:2609.26911v1 Announce Type: new Abstract: A single locally plausible tool call can derail an otherwise successful agent trajectory. Suspicion alone does not justify intervention, because the rep

When Context Misleads: In-context Learning with Jurisdiction in Large Language Models

arXiv:2609.27603v1 Announce Type: cross Abstract: In-Context Learning (ICL) has become a cornerstone of modern LLM deployment. However, existing ICL post-training methods have a critical blind spot: t

Connecting the Dots in Agentic AI Security: A Cross-Dimensional Threat Taxonomy, Evaluation Maturity, and Open Challenges

arXiv:2609.23894v1 Announce Type: cross Abstract: Agentic AI extends LLM security beyond generated content to persistent state, autonomous actions, tool use, and interactions with humans and other age

Custom Node for Extract Prompt & Target Aspect Ratio/Dimensions from Qwen Image 2.1 PE

model-releasesr-stablediffusion

Hey everyone! I've been experimenting with Qwen Image 2.1 Prompt Enhancer (PE) lately and noticed an interesting detail in Qwen's official system prompts. The PE models output structured JSON containi

EDGEGEN: Improving Tool-Calling Agents Beyond Happy Paths with Synthetic Edge Case Generation

arXiv:2609.24115v1 Announce Type: new Abstract: Tool-calling LLM agents are increasingly deployed in enterprise applications. However, effective evaluation and optimization require high-quality, diver

FinFIRST: Benchmarking Search Agents for Financial Information Retrieval, Sourcing and Traceability

arXiv:2609.25192v1 Announce Type: new Abstract: Financial search is a highly demanding task for LLM agents, requiring not only a correct final answer but also temporally valid information retrieval, a

How Many Pixels Is a Digit Worth? Place-Aware Coordinate Entropy for GUI Agent Confidence Estimation

arXiv:2609.24277v1 Announce Type: new Abstract: GUI agents predict click coordinates as digit-token sequences, but standard text-LLM confidence estimation methods rank correct clicks from wrong ones o

Local Ai

MAGIC: Mixed-Granularity Agent Graphs via Incremental Construction with Dense-Reward Reinforcement Learning

arXiv:2609.26667v1 Announce Type: new Abstract: Collaboration topology shapes both the performance and execution cost of LLM-based multi-agent systems. Because tasks differ in complexity and required

local-aiarxiv-cs-lg

MoM: Memory of Memory

arXiv:2609.25054v1 Announce Type: new Abstract: For a long-horizon LLM agent, the memory question is not what was once recorded but what currently holds. Most designs answer it only indirectly: every

agentsarxiv-cs-cl

PACT: From Credit Assignment to Critic Alignment

arXiv:2609.26355v1 Announce Type: new Abstract: Reinforcement learning has become a central component of large language model (LLM) post-training, yet token-level credit lacks a generally accepted mat

safetyarxiv-cs-lg

PERSONAWEAVER: Controllable Diversity Beyond Conventional Archetypes in Procedural Character Generation

arXiv:2609.26629v1 Announce Type: new Abstract: Procedural character generation aims to populate games, simulations, and other virtual worlds with diverse characters. Large language models (LLMs) offe

researcharxiv-cs-cl

PINNForge: Execution-Grounded Evolutionary Design of Physics-Informed Neural Networks for PDE Solving via Large Language Models

arXiv:2609.23023v1 Announce Type: new Abstract: Physics-informed neural networks (PINNs) require coordinated choices over network representation, sampling, loss construction, and optimization, while e

researcharxiv-cs-ai

Local Ai

Self-Healing Harness for Runtime Oversight of Agent Self-Modification

arXiv:2609.24130v1 Announce Type: new Abstract: LLM agents can change their own future behavior, raising a basic control question of which self-generated changes should be allowed to persist. We formu

local-aiarxiv-cs-ai

SelfOp: An Optimization Algorithm for Self-Improving Security Agents

arXiv:2609.22792v1 Announce Type: cross Abstract: LLM agents are increasingly used for security tasks: vulnerability discovery, exploit reproduction, and patch generation. Improving them at the model

SPECTRA: Adaptive Execution of Speculative Decoding on a Runtime-Reconfigurable Tiled Architecture

arXiv:2609.24847v1 Announce Type: cross Abstract: LLM inference on edge devices is constrained by computational and memory resources, making efficient autoregressive decoding challenging. Speculative

researcharxiv-cs-ai

stuntd: a local Jev-compatible server on Laya that learns from your own traffic (no API key needed)

model-releasesr-localllama

What it is. stuntd is a local proxy for typed LLM decisions: choice, yes/no, score. Your app keeps calling its provider through it. stuntd records the decisions, trains a small head on top of Laya (th

Toolcompass: Guiding Tool Trialing, Not Suppressing It

arXiv:2609.25678v1 Announce Type: cross Abstract: Large language model (LLM) agents must generalize from tools seen during training to unseen tools at deployment. A key challenge is tool trialing, i.e

researcharxiv-cs-lg

Transcribe, Translate, and Optimize: Joint Reward Learning for Speech Translation

arXiv:2609.26536v1 Announce Type: new Abstract: In LLM-based speech translation, transcription-based chain-of-thought (CoT) suffers from a mismatch between reference transcripts used in supervised fin

safetyarxiv-cs-cl

VeriSoftBench: Repository-Scale Formal Verification Benchmarks for Lean

arXiv:2602.18307v2 Announce Type: replace-cross Abstract: Large language models have achieved striking results in interactive theorem proving, particularly in Lean. However, most benchmarks for LLM-ba

Which small model works best with Ollama on a Raspberry Pi 5?

I've been comparing and working on Gemma 4, Phi-4-Mini, and Qwen3-4B for local LLM use on an 8 GB Raspberry Pi 5. The results were indeed close, but each model has a different strength: Gemma 4 E2B: a

model-releasesr-ollama

Zero-Trust Authorization and Discovery for Enterprise MCP

arXiv:2609.22573v1 Announce Type: cross Abstract: LLM agents translate natural-language context, which may include attacker-controlled text, into privileged tool calls, so authorization must remain ef

Acceptance-Aware Draft Model Training for Speculative Decoding

arXiv:2609.24150v1 Announce Type: new Abstract: Speculative decoding accelerates large language model (LLM) inference by using a lightweight draft model to generate multiple candidate tokens that are

researcharxiv-cs-lg

Adapting Tree-Structured Speculative Decoding to DeepSeek-V4 for Efficient Inference

arXiv:2609.24698v1 Announce Type: new Abstract: Repeated execution of the target model during autoregressive decoding is a major source of LLM inference latency. Unlike linear speculation, which follo

An Empirical Cost Attribution of Context-Compression Gateways in Multi-Turn Coding Agents

arXiv:2609.22114v1 Announce Type: new Abstract: Context compression is widely proposed as a way to cut the token bill of LLM coding agents, and public benchmarks report that aggressive compression pre

Beyond Single-Model Injection: A Threat Model and Defense Architecture for Prompt Injection in Multi-Agent Systems

arXiv:2609.22949v1 Announce Type: cross Abstract: Existing prompt injection research focuses on single-model chatbot scenarios, where an attacker manipulates one LLM through crafted input. Multi-agent

agentsarxiv-cs-cl

Cost of a local set-up for coding?

model-releasesr-localllama

I have to say that I am quite impressed by recent progress in LLM use for coding. I've never really bothered setting up a local model. However now that I went trough some real $$ in Claude / gpt token

Does the Truthfulness Signal Survive Code-Mixing? Probing Hidden States for Hallucination Detection in Hinglish

arXiv:2609.22138v1 Announce Type: new Abstract: Hidden-state hallucination probing - training a linear classifier on an LLM's internal activations to detect whether a generated answer is faithful to t

EvalMem: An Operation-Level Diagnostic Framework for Long-Term Memory Systems

arXiv:2609.22231v1 Announce Type: new Abstract: Long-horizon interactions with LLM-based assistants require memory systems that preserve and update user states, preferences, and interaction histories.

agentsarxiv-cs-cl

Evaluating Decision Models for Text Annotation in Computational Social Science

arXiv:2609.24574v1 Announce Type: new Abstract: Computational social science increasingly relies on large language models for text annotation, and the validity of published findings now rests on the l

FLARE: A Full-Lifecycle Dense Supervision Paradigm for Long-Horizon Coding Agents via Generative Reward Model

arXiv:2609.23808v1 Announce Type: new Abstract: While test-time scaling enhances Large Language Model (LLM) agents in long-horizon software engineering (SWE), sparse binary rewards (Pass/Fail) create

safetyarxiv-cs-cl

FRAMES: Failure Recovery And Monitoring of Embodied Skills for Humanoid Loco-Manipulation

arXiv:2609.22538v1 Announce Type: new Abstract: Large language model (LLM) planners can decompose natural-language instructions and select reusable robot skills, but choosing the correct skill does no

agentsarxiv-cs-ro

From Documented Strengths to Force Limits: Material-Informed Robotic Insertion for Construction Assembly

model-releasesarxiv-cs-ro

arXiv:2609.22609v1 Announce Type: new Abstract: Insertion is a fundamental operation in robotic construction assembly, where variations in material properties and assembly conditions make it difficult

Applications

Heterogeneous Robot Collaboration in Unstructured Environments with Grounded Generative Intelligence

arXiv:2510.26915v2 Announce Type: replace Abstract: While heterogeneous teams have typically been designed for well-specified missions with known semantics, generative intelligence, i.e., large langua

applicationsarxiv-cs-ro

Measuring the Checker: Mutation Analysis for GPU-Kernel Benchmark Oracles

model-releasesarxiv-cs-lg

arXiv:2609.22220v1 Announce Type: new Abstract: Benchmarks for LLM-generated GPU kernels decide correctness with a few random inputs and a loose floating-point tolerance, and their verdicts now feed l

OSCAR: Order-aware Scoring and Calibration for AI Rankings

model-releasesarxiv-cs-lg

arXiv:2609.24128v1 Announce Type: cross Abstract: Judge-specific sensitivity is useful for aggregating pairwise LLM evaluations, but its interpretation depends on which systematic presentation effects

Pretrained Persona Mixture Models and Tandem Models for Human Simulation

arXiv:2609.22607v1 Announce Type: new Abstract: We argue here that the current dominant practice in LLM human simulation: prompting instruction-tuned assistant language models to role-play personas, i

researcharxiv-cs-cl

RRSI: Regularized Recursive Self-Improvement of Agent Harnesses

arXiv:2609.24972v1 Announce Type: cross Abstract: An LLM agent's capability is largely magnified by its harness, namely the prompts, control flow, tooling, memory, and context management surrounding t