Search: “open-weights”

RusFinChain: A Russian Benchmark for Verifiable Chain-of-Thought Reasoning in Finance with Fuzzy-Aligned Evaluation

arXiv:2607.01388v1 Announce Type: new Abstract: Multi-step symbolic reasoning is essential for robust financial analysis, yet most benchmarks neglect intermediate reasoning steps. FINCHAIN introduced

3 Jul 2026

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Research

Will Scaling Improve Social Simulation with LLMs?

DGX agent

arXiv:2607.02464v1 Announce Type: new Abstract: Large Language Model (LLM) social simulations are a promising research method, but they are not yet faithful enough to be adopted widely. In this work,

researcharxiv-cs-cl

3 Jul 2026

AGC-Bench: Measuring Artificial General Creativity

arXiv:2607.01152v1 Announce Type: new Abstract: Creativity research has debated whether creativity is domain-specific (e.g., visual, writing, science), and if it is psychometrically separable from gen

An LLM-Based Framework for Intent-Driven Network Topology Design

arXiv:2607.00292v1 Announce Type: cross Abstract: Designing deployable and resilient network topologies from natural language requirements remains a challenging problem in network automation. This wor

AutoMem: Automated Learning of Memory as a Cognitive Skill

arXiv:2607.01224v1 Announce Type: new Abstract: Memory expertise is a learned skill: knowing what to encode, when to retrieve, and how to organize knowledge--a capacity known in cognitive science as m

'Don't Say It!': Constraints, Compliance, and Communication when Language Models Play Taboo

arXiv:2607.00601v1 Announce Type: new Abstract: The game of Taboo requires describing a target word without using a set of forbidden words, so that other players can guess it. This deceptively simple

researcharxiv-cs-cl

Rethinking Visual Privacy: A Compositional Privacy Risk Framework for Severity Assessment with VLMs

arXiv:2603.21573v2 Announce Type: replace Abstract: Existing visual privacy benchmarks largely treat privacy as a binary property, labeling images as private or non-private based on visible sensitive

researcharxiv-cs-cv

WorkBench Revisited: Workplace Agents Two Years On

arXiv:2606.13715v2 Announce Type: replace Abstract: The best agent on WorkBench in March 2024, GPT-4, completed just 43% of tasks. We revisit the benchmark in June 2026 and find that the best agent to

Applications

AI-Generated PowerShell Malware: An Experimental Framework and Dataset

arXiv:2606.30819v1 Announce Type: cross Abstract: Generative AI has emerged as a significant cybersecurity threat, with several recent attack campaigns leveraging LLMs to generate code for malicious p

applicationsarxiv-cs-ai

DetPO: In-Context Learning with Multi-Modal LLMs for Few-Shot Object Detection

arXiv:2603.23455v2 Announce Type: replace Abstract: Multi-Modal LLMs (MLLMs) demonstrate strong visual grounding capabilities on popular object detection benchmarks like OdinW-13 and RefCOCO. However,

researcharxiv-cs-cv

Indi-RomCoM: Code-Mixed Benchmark for Evaluating LLMs on Romanized Indic-English Instructions

arXiv:2606.30790v1 Announce Type: cross Abstract: Romanized Code Mixing (RCM), where bilingual speakers fluidly blend local languages with English in Roman script, has emerged as the dominant form of

Safety

QVal: Cheaply Evaluating Dense Supervision Signals for Long-Horizon LLM Agents

arXiv:2606.32034v1 Announce Type: cross Abstract: LLM agents increasingly act over long horizons, where a single trajectory can contain hundreds or thousands of actions. In these settings, outcome-onl

safetyarxiv-cs-ai

RoPoLL: Robust Panel of LLM Judges

arXiv:2606.30931v1 Announce Type: new Abstract: The LLM Jury, a Panel of LLM Evaluators (PoLL) reporting consensus scores, has become a practical alternative to single-judge LLM evaluation, yet its st

What Drives Interactive Improvement from Feedback?

arXiv:2606.30774v1 Announce Type: new Abstract: We study when natural-language feedback produces improvement beyond the gains obtainable from repeated attempts alone. In multi-turn language agent sett

agentsarxiv-cs-ai

When the Database Fails: Prompting LLM Dialogue Agents for Safe Recovery in Task-Oriented Dialogue

arXiv:2606.31307v1 Announce Type: new Abstract: Large language models used in task-oriented dialogue often produce fluent but unsafe responses when backend database calls fail, return empty results, o

A Comparative Study on Affective Cues in Text Embeddings Across Psychological Emotion Theories

arXiv:2606.29068v1 Announce Type: cross Abstract: Text encoders are known for their utility in natural language processing, as they are able to efficiently compress inputs into dense vectors while pre

Safety

Can LLMs Reliably Self-Report Adversarial Prefills, and How?

arXiv:2606.23671v2 Announce Type: replace Abstract: Prior work shows that large language models (LLMs) exhibit introspective capability on benign tasks. We extend the question to safety contexts and e

safetyarxiv-cs-cl

Safety

Can MLLMs Critique Like Humans? Evaluating Open-Ended Aesthetic Reasoning in Multimodal Large Language Models

arXiv:2606.29689v1 Announce Type: new Abstract: Open-ended aesthetic critique is a challenge for multimodal large language models (MLLMs): unlike multiple-choice aesthetic benchmarks, it has no single

safetyarxiv-cs-cl

Collective cooperation without individual fidelity in LLM agents

arXiv:2606.30454v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly used as agents in simulations of social systems, yet it remains unclear when their behavior can be inter

Diff-Based Code Corruption using LLMs for Large-Scale Bugfix Benchmarking

arXiv:2606.29088v1 Announce Type: cross Abstract: There are various benchmarks to evaluate bugfixing capabilities of Large Language Models. However, most widespread benchmarks do not fully reflect rea

EMPATH: A Multilingual Auditor-Judge Benchmark for Safety Evaluation of Emotional-Support Chatbots

arXiv:2606.30256v1 Announce Type: new Abstract: Safety benchmarks often buy scalability by fixing the prompt, the language, and the turn structure. For emotional-support chatbots, that bargain hides p

Enhancing Automatic Chord Recognition via Pseudo-Labeling and Knowledge Distillation

arXiv:2602.19778v4 Announce Type: replace-cross Abstract: Automatic Chord Recognition (ACR) is constrained by the scarcity of aligned chord labels, as well-aligned annotations are costly to acquire. A

researcharxiv-cs-lg

From Accuracy to Visual Dependence: Auditing and Filtering Modality Collapse in Traffic VideoQA

model-releasesarxiv-cs-cv

arXiv:2606.30220v1 Announce Type: new Abstract: High benchmark accuracy does not guarantee genuine use of visual evidence. We study this problem in traffic accident Video Question Answering (VideoQA),

Hierarchical Experimentalist Agents

arXiv:2606.29315v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used to take actions in the real world and support human decision-making, yet most agents rely on parametr

How to Train Your Long-Context Visual Document Model

arXiv:2602.15257v3 Announce Type: replace-cross Abstract: We present the first comprehensive, large-scale study of training long-context vision language models up to 344K context, targeting long-docum

SADL: What to Ignore? A Benchmark for Subject-Aware Distractor Localization

model-releasesarxiv-cs-cv

arXiv:2606.30393v1 Announce Type: new Abstract: Photographs frequently contain visual distractors besides foregrounds and backgrounds of the intended subject, competing for attention and weakening com

SAKE: Software Architectural Knowledge Evaluation Benchmark for Large Language Models

arXiv:2606.29520v1 Announce Type: cross Abstract: Large Language Models (LLMs) are increasingly used as assistants across the software development lifecycle, yet their ability to reason about software

SWE-INTERACT: Reimagining SWE Benchmarks as User-Driven Long-Horizon Coding Sessions

arXiv:2606.30573v1 Announce Type: new Abstract: We introduce SWE-Interact, a new testbed for evaluating coding agents on multi-turn, interactive, user-driven software engineering tasks. Existing front

agentsarxiv-cs-lg

The Complexity Ceiling Benchmark: A Multi-Domain Evaluation of Sequential Reasoning Under Depth Scaling

arXiv:2606.29278v1 Announce Type: new Abstract: We introduce the Complexity Ceiling Benchmark (CCB), a controlled evaluation of how language-model reasoning decays as the number of required sequential

The Emergence of Autonomous Penetration Capabilities in Large Language Model-Powered AI Systems

arXiv:2606.13079v2 Announce Type: replace-cross Abstract: Nowadays, the autonomous execution of cyberattacks capable of causing substantial real-world harm is widely regarded as one of the critical re

agentsarxiv-cs-ai

Local Ai

The strength of clinical evidence is recoverable from language model representations but not from their stated grades

arXiv:2606.29034v1 Announce Type: cross Abstract: Large language models (LLMs) increasingly summarize clinical evidence, where a claim's weight depends on how strongly it is supported. Yet these model

local-aiarxiv-cs-ai

When Can Conformal Risk Control Certify LLM Outputs? Bounds, Impossibility, and Adaptation for Structured Generation

arXiv:2606.29054v1 Announce Type: new Abstract: Large language models (LLMs) deployed for structured generation (NER, JSON extraction, QA, and classification) lack formal reliability guarantees, and s

researcharxiv-cs-lg

Whose Side Is Your Agent On? Multi-Party Principal Loyalty in LLM Agents

arXiv:2606.30383v1 Announce Type: new Abstract: A rapidly growing class of LLM agents is multi-party: the agent acts for a principal (who briefs it, sends follow-ups, and receives results) while also

DMind Benchmark: Toward a Holistic Assessment of LLM Capabilities across the Web3 Domain

arXiv:2504.16116v4 Announce Type: replace-cross Abstract: The Web3 ecosystem, underpinned by cryptographic primitives and decentralized consensus, represents a high-stakes environment where software v

29 Jun 2026

Local Ai

Enhancing Numerical Prediction in LLMs via Smooth MMD Alignment

arXiv:2606.27731v1 Announce Type: cross Abstract: Despite their strong general capabilities, large language models (LLMs) often remain unreliable when outputs must be numerically precise. A key reason

local-aiarxiv-cs-ai

29 Jun 2026

Formalizing Latent Thoughts: Four Axioms of Thought Representation in LLMs

arXiv:2606.27378v1 Announce Type: new Abstract: We introduce an axiomatic evaluation framework for latent thought representations in LLMs, comprising metrics that are independent of downstream benchma

29 Jun 2026

Adaptive Evaluation of Out-of-Band Defenses Against Prompt Injection in LLM Agents

arXiv:2606.26479v1 Announce Type: cross Abstract: Recent work (2024 to 2026) has converged on a strategy for defending tool-using LLM agents against indirect prompt injection: rather than training the

Advancing Omnimodal Embodied Agents from Isolated Skills to Everyday Physical Autonomy

arXiv:2606.27251v1 Announce Type: cross Abstract: Building persistent embodied agents in unstructured environments demands unified orchestration of heterogeneous tools spanning both cyber (APIs, IoT)

agentsarxiv-cs-ai

Can Large Language Models Reliably Code Qualitative Humanitarian Data? A Benchmark Study Against Human Expert Adjudication

model-releasesarxiv-cs-lg

arXiv:2606.26541v1 Announce Type: new Abstract: Data from affected populations are crucial for informing humanitarian response, but their value depends on timely and consistent interpretation of nuanc

Comparing BERT Sentence-Pair Classification and Few-Shot LLM Prompting for Detecting Threat and Solution Framing in German Climate News

arXiv:2606.26489v1 Announce Type: new Abstract: News media play a central role in shaping public perceptions of climate change, and whether coverage emphasizes threats or solutions has measurable effe

Jailbreaking for the Average Jane: Choosing Optimal Jailbreaks via Bandit Algorithms for Automatically Enhanced Queries

arXiv:2606.26936v1 Announce Type: cross Abstract: With a profusion of jailbreaks for LLMs now widely known, a growing concern is that non-expert malicious actors ('the average Jane') could elicit acti

Applications

Mapping Political-Elite Networks in Europe with a Multilingual Joint Entity-Relation Extraction Pipeline

arXiv:2606.27347v1 Announce Type: new Abstract: Whether political elites organise into rent-seeking coalitions that capture public resources or civic networks that sustain governance is a central ques

applicationsarxiv-cs-cl

Nemotron-TwoTower: Diffusion Language Modeling with Pretrained Autoregressive Context

arXiv:2606.26493v1 Announce Type: new Abstract: Diffusion language models offer a promising alternative to autoregressive models due to their potential for parallel and iterative generation. However,

SocialPersona: Benchmarking Personalized Profiling and Response with Multimodal Social-Media Context

arXiv:2606.26654v1 Announce Type: new Abstract: Personalized language-model assistants are often evaluated through a memory lens: can a model recall preferences users have explicitly stated in dialogu

Where Do Models Find Happiness? Emotion Vectors in Open-Source LLMs

arXiv:2606.26987v1 Announce Type: cross Abstract: Recent work identified emotion vectors in Claude Sonnet 4.5, which are internal representations that encode emotion concepts, causally influence behav

Beyond Visual Forensics: Auditing Multimodal Robustness for Synthetic Medical Image Detection

model-releasesarxiv-cs-cv

arXiv:2606.25375v1 Announce Type: new Abstract: With the rapid adoption of generative AI, synthetic medical images pose growing risks, including diagnostic deception and insurance fraud. Although prio

25 Jun 2026

Brevity is the Soul of Inference Efficiency: Inducing Concision in VLMs via Data Curation

arXiv:2606.25432v1 Announce Type: cross Abstract: Inference efficiency is typically pursued by shrinking the model: distillation, pruning, quantization, and sparse routing each lower per-token cost wh

researcharxiv-cs-cv

25 Jun 2026

Do Thinking Tokens Help with Safety?

arXiv:2606.25013v1 Announce Type: cross Abstract: Today's reasoning models use thinking tokens to attain stronger performance on benchmarks than their instruction-tuned counterparts. It is also genera