Search: “open-weights”

Opir: Efficient Multi-Task Safety Classification for Toxicity, Jailbreaks, Hate Speech, and Harmful Content

arXiv:2605.29659v1 Announce Type: cross Abstract: Real-time safety filtering for large language model (LLM) applications requires classifiers that can detect unsafe prompts, toxic language, jailbreak

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Model Releases

REPOT: Recoverable Program-of-Thought via Checkpoint Repair

DGX agent

arXiv:2605.30052v1 Announce Type: cross Abstract: One-shot Program-of-Thought (PoT) emits a Python program that prints a primitive-action plan; a single invalid action silently invalidates the traject

model-releasesarxiv-cs-ai

29 May 2026

SciIntBench: Measuring LLM Compliance with Research Integrity Norms Under Adversarial Framing

arXiv:2605.29468v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly used to support scientific work, but it is unclear whether they uphold responsible conduct of research (

Slogans or Stance? A Label-Light Diagnostic for Entrepreneurial-Discourse Measurement on Chinese SOE Speeches

arXiv:2605.29188v1 Announce Type: new Abstract: Dictionary methods, topic models, and embedding-similarity scorers are widely used in CSS and management research to measure constructs such as 'entrepr

When the Same Coefficients Reach Different Places: Asymmetric Realizability in Transplanting Tokenizers across Large Language Models

arXiv:2601.00065v3 Announce Type: replace-cross Abstract: Tokenizer transplant in cross-vocabulary model composition reconstructs donor-only embedding rows as weighted combinations over shared lexical

Argument Quality Assessment with Large Language Models: A Pairwise Bradley-Terry Approach

arXiv:2605.28313v1 Announce Type: new Abstract: Large Language Models (LLMs) have demonstrated remarkable capabilities in tasks related to reasoning and judgment. However, assessing the quality of arg

Automating Formal Verification with Agent-Guided Tree Search

model-releasesarxiv-cs-lg

arXiv:2605.27485v1 Announce Type: cross Abstract: Formal verification offers a path to provably correct software, but writing verified code remains expensive enough that the technique is rarely used i

BenGER: Benchmarking LLM Systems on Subsumption-Based Legal Reasoning in German Law

arXiv:2605.28183v1 Announce Type: cross Abstract: We introduce the BenGER (Benchmark for German Law) dataset for evaluating LLM systems on subsumption-based legal reasoning in German law. The BenGER d

Disentangling Language Roles in Multilingual LLM Task Execution

arXiv:2605.27649v1 Announce Type: new Abstract: Multilingual LLMs are increasingly used when instruction, source content, and required response languages do not coincide. Existing benchmarks have expa

Escape the Language Prior: Mitigating Late-Stage Modality Collapse in Audio Reasoning via Modality-Aware Policy Optimization

arXiv:2605.27741v1 Announce Type: new Abstract: Audio and omni-modal large language models exhibit impressive cross-modal reasoning capabilities. However, applying standard reinforcement learning post

safetyarxiv-cs-cl

Geometry of Human Perceptual Domains Emerges Transiently in LLM Representations

arXiv:2605.27970v1 Announce Type: new Abstract: While large language models (LLMs) are trained purely on textual data, prior work has shown that their internal representations can exhibit rich geometr

safetyarxiv-cs-ai

Agents

Large Language Models Approach Expert Pedagogical Quality in Math Tutoring but Differ in Instructional and Linguistic Profiles

arXiv:2512.20780v3 Announce Type: replace Abstract: Recent work has explored the use of large language models (LLMs) to generate tutoring responses in mathematics, yet it remains unclear how closely t

agentsarxiv-cs-cl

Position: Retire the 'Positive Backdoor' Label -- Secret Alignment Requires Strict and Systematic Evaluation

arXiv:2605.28597v1 Announce Type: cross Abstract: This position paper argues that the AI/ML community should stop overclaiming and retire the label 'positive backdoor,' and instead treat trigger-activ

safetyarxiv-cs-ai

Prefix-Safe Bayesian Belief Tracking for LLM Reasoning Reliability:Separating Calibration from Ranking

arXiv:2605.27712v1 Announce Type: new Abstract: Long reasoning traces need reliability estimates before final answers are known. We study prefix-conditioned eventual-success estimation, P(y=1 mid o_{1

Local Ai

Reading or Guessing? Visual Grounding Failures of Vision-Language Models for OCR in Ancient Greek Editions

arXiv:2605.27750v1 Announce Type: cross Abstract: Recent work has shown that Vision-Language Models (VLMs) used for optical character recognition (OCR) can generate plausible but visually unsupported

local-aiarxiv-cs-ai

Agents

ResearchMath-14K: Scaling Research-Level Mathematics via Agents

arXiv:2605.28003v1 Announce Type: new Abstract: The frontier of mathematics is defined by problems whose solutions are not yet known, yet it remains unclear whether language models can meaningfully en

agentsarxiv-cs-cl

Soro: A Lightweight Foundation Model and Chatbot for Tajik

arXiv:2605.27379v1 Announce Type: new Abstract: We present Soro, a family of Tajik-specialized conversational large language models (LLMs) designed for real-world deployment under tight compute and co

The Harder Text Embedding Benchmark (HTEB): Beyond One-dimensional Static Robustness

arXiv:2605.28190v1 Announce Type: new Abstract: Embedding benchmarks like MTEB report a single score per model, implicitly treating robustness as a static, scalar property. We argue that embedding rob

The Importance of Being Statistically Earnest: A Critical Re-evaluation of GSM-Symbolic

arXiv:2605.28700v1 Announce Type: new Abstract: The GSM-Symbolic benchmark (Mirzadeh et al., 2025) reported consistent performance drops across 25 Large Language Models (LLMs) when tested on template-

Hardware

Today, we're releasing LFM2.5-8B-A1B, a device-optimized model designed to power real-life applications on phones, laptops, PCs, robots, and…

Today, we're releasing LFM2.5-8B-A1B, a device-optimized model designed to power real-life applications on phones, laptops, PCs, robots, and fast & lightweight server-side use-cases. > 8B MoE, 1.5B ac

hardwareclem-delangue--x

AirCast-SR: A Foundation Model for Kilometer-Scale Atmospheric Super-Resolution via Latent Consistency Diffusion

arXiv:2605.26130v1 Announce Type: new Abstract: Operational weather prediction at kilometer scales remains computationally prohibitive for traditional numerical weather prediction (NWP) models, limiti

safetyarxiv-cs-lg

LLM-guided Hierarchical Search for End-to-end Reasoning Intensive Retrieval

model-releasesarxiv-cs-lg

arXiv:2510.13217v2 Announce Type: replace-cross Abstract: Search systems are increasingly used for reasoning-intensive queries, where what makes a document relevant requires understanding or reasoning

Lost in Sampling: Assessing Lexical Reachability in LLMs via the Word Coverage Score (WCS)

arXiv:2605.27268v1 Announce Type: cross Abstract: Modern Large Language Models (LLMs) are often criticized for producing repetitive and homogeneous text, despite possessing vast latent vocabularies. W

SEC-bench Pro: Can Language Models Solve Long-Horizon Software Security Tasks?

model-releasesarxiv-cs-lg

arXiv:2605.26548v1 Announce Type: cross Abstract: Large language models (LLMs) now support automated software security tasks, including vulnerability discovery and proof-of-concept (PoC) generation. E

Uncertainty-Aware Budget Allocation for Adaptive Test-Time Reasoning

arXiv:2605.26849v1 Announce Type: new Abstract: Sampling multiple responses improves language model reasoning, but uniform compute allocation is inefficient: easy questions are over-sampled while hard

AstroMind: A High-Fidelity Benchmark for Spacecraft Behavior Reasoning Based on Large Language Models

arXiv:2605.24573v1 Announce Type: new Abstract: Understanding why a spacecraft maneuvers -- rather than simply that it did -- is an increasingly important problem for space domain awareness as Earth o

CUNY at CLPsych 2026: A Pipeline Approach to Classification and Summarization of Mental Health Changes

arXiv:2605.24164v1 Announce Type: new Abstract: We describe our submission to the CLPsych~2026 Shared Task on capturing and characterizing mental health changes through social media timeline dynamics.

EPPC-OASIS: Ontology-Aware Adaptation and Structured Inference Refinement for Electronic Patient-Provider Communication Mining in Secure Messages

arXiv:2605.24172v1 Announce Type: new Abstract: Secure patient-provider messages contain clinically important communication behaviors that are difficult to characterize manually at scale. The Electron

safetyarxiv-cs-ai

Explaining Too Much? Understanding How Large Language Model Reasoning Traces Influence Performance and Metacognition

arXiv:2605.25856v1 Announce Type: cross Abstract: Large Language Model interfaces are increasingly verbose, exposing intermediate reasoning traces alongside final answers. Traces are framed as transpa

Local Ai

False Fixed Points: Kantian Feedback, Stable Miscalibration, and Representational Compression in LLMs

arXiv:2510.14925v4 Announce Type: replace Abstract: High-confidence errors in large language models are often treated as fragile failures. We study an alternative: some errors may be false fixed point

local-aiarxiv-cs-ai

Geo-Expert: Towards Expert-Level Geological Reasoning via Parameter-Efficient Fine-Tuning

arXiv:2605.24844v1 Announce Type: new Abstract: While general-purpose Large Language Models (LLMs) applied to Geology often hallucinate when reasoning about subsurface structures and deep-time evoluti

Is Inference Mediated by Distinct Semantic Structures in LLMs? A Mechanistic Interpretation

arXiv:2605.25520v1 Announce Type: new Abstract: Predicting a label correctly does not necessarily require representing the operation that produces it. Transformer representations are known to carry la

Large Language Model Selection with Limited Annotations

arXiv:2605.24981v1 Announce Type: new Abstract: Choosing a Large Language Model (LLM) for a given task requires comparing many strong candidates, yet standard evaluation relies on costly annotations o

Tutorials

Leveraging Spreading Activation for Improved Document Retrieval in Knowledge-Graph-Based RAG Systems

arXiv:2512.15922v3 Announce Type: replace Abstract: Despite initial successes and a variety of architectures, retrieval-augmented generation systems still struggle to reliably retrieve and connect the

tutorialsarxiv-cs-ai

Hardware

Paris 2.0: A Decentralized Diffusion Model for Video Generation

arXiv:2605.26064v1 Announce Type: cross Abstract: We present Paris 2.0, the first video generation model pre-trained through decentralized computation. Its training recipe builds upon Paris 1.0 (arXiv

hardwarearxiv-cs-lg

Pragmatic Reasoning improves LLM Code Generation

arXiv:2502.15835v5 Announce Type: replace-cross Abstract: Pragmatic reasoning helps interlocutors infer intended meaning from ambiguous or underspecified messages by considering shared context and cou

PromptAudit: Auditing Prompt Sensitivity in LLM-Based Vulnerability Detection

arXiv:2605.24171v1 Announce Type: cross Abstract: Large language models are increasingly used for vulnerability detection, yet their reliability under different prompt formulations remains uncharacter

QUEST: Training Frontier Deep Research Agents with Fully Synthetic Tasks

arXiv:2605.24218v1 Announce Type: new Abstract: Deep research agents extend the role of search engines from retrieving keyword-matched pages to synthesizing knowledge, fundamentally changing how human

Residual Drift Dominates Contradiction in Multi-Turn Constraint Reasoning

arXiv:2605.23940v1 Announce Type: new Abstract: How do multi-turn reasoning systems fail? The expected answer is logical contradiction, in which the system's maintained state becomes unsatisfiable. We

Tiny Brains, Giant Impact: Uncovering the Keystone Neurons of LLM with Just a Few Prompts

arXiv:2605.24846v1 Announce Type: cross Abstract: Large language models (LLMs) display strong comprehensive abilities, yet the internal mechanisms that support these behaviors remain insufficiently un

An open source model has returned to #1 on the 3D Design leaderboard by Design Arena. Kimi K2.6 has reached the top of the leaderboard for 3…

model-releaseskimi-moonshot--x

An open source model has returned to #1 on the 3D Design leaderboard by Design Arena. Kimi K2.6 has reached the top of the leaderboard for 3D Design, ahead of models 10X more expensive like Opus 4.7 b

BURMESE-SAN: Burmese NLP Benchmark for Evaluating Large Language Models

arXiv:2602.18788v3 Announce Type: replace Abstract: We introduce BURMESE-SAN, the first holistic benchmark that systematically evaluates large language models (LLMs) for Burmese across three core NLP

GENSTRAT: Toward a Science of Strategic Reasoning in Large Language Models

arXiv:2605.23238v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly deployed as economic agents in marketplaces, auctions, and bidding settings. Anticipating their behavior i

It's the humans, not the data: Geopolitical bias in LLMs originates in post-training, amplified by the language of the prompt

arXiv:2605.23825v1 Announce Type: cross Abstract: It has generally been assumed that geopolitical bias in language models originates from the training data used during the pre-training phase. We teste

PoisonForge: Task-Level Targeted Poisoning Benchmark for Instruction-Tuned LLMs

arXiv:2605.23168v1 Announce Type: cross Abstract: When practitioners fine-tune LLMs on unvetted datasets, an adversary can exploit the data supply chain through task-level poisoning: inserting a small

Polite on the Surface, Wrong in Practice: A Curated Dataset for Fixing Honorific Failures in Multilingual Bangla Generation

arXiv:2605.22487v1 Announce Type: new Abstract: Recent advances in Multilingual Large Language Models (MLLMs) have significantly enhanced cross-lingual conversational capabilities, yet modeling cultur

22 May 2026

Seizure-Semiology-Suite (S3): A Clinically Multimodal Dataset, Benchmark, and Models for Seizure Semiology Understanding

model-releasesarxiv-cs-cv

arXiv:2605.21852v1 Announce Type: new Abstract: While Multimodal Large Language Models (MLLMs) have demonstrated remarkable proficiency in general video understanding, their capacity to interpret invo

22 May 2026

A Systematic Comparison between Extractive Self-Explanations and Human Rationales in Text Classification