Search: “evaluation”

Data World: Scaling Data Environments from Scratch

arXiv:2610.21656v1 Announce Type: new Abstract: Data agents use code to process heterogeneous files and derive analytical insights. Scaling their training faces costly environment curation, privacy an

Evolving Together, Failing Apart: The Skill-Policy Co-evolution Trap in Self-Improving Agents

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

DGX agent

arXiv:2610.20845v1 Announce Type: new Abstract: Skill-policy co-evolution, in which an agent's policy model and its skill bank improve each other during training, is a promising paradigm for self-impr

safetyarxiv-cs-ai

16 Oct 2026

Exploring the Design Space of Permutation Diffusion LLMs

arXiv:2610.20635v1 Announce Type: cross Abstract: Permutation generation with language models requires both text-conditioned ranking and a global one-to-one assignment between candidates and output po

FlowBudget: Training-Free Sampling-Budget Estimation for Flow Matching Models

arXiv:2610.20636v1 Announce Type: cross Abstract: Sampling from continuous flow-based generators such as flow matching models requires numerical ODE or SDE integration, making inference cost depend on

researcharxiv-cs-cv

IDfRA: Self-Verification for Iterative Design in Robotic Assembly

arXiv:2509.16998v2 Announce Type: replace Abstract: Design for Robotic Assembly (DfRA) remains largely dependent on manual planning and heuristic simulation, limiting scalability and robustness in com

applicationsarxiv-cs-ro

International Multi-Centre Validation of Unsupervised Domain Adaptation for Precise Discrimination between Normal and Abnormal Retinal Development

arXiv:2610.21719v1 Announce Type: new Abstract: Arrested retinal development can cause childhood visual impairment, but grading optical coherence tomography scans is difficult to standardise across de

researcharxiv-cs-cv

Language Is a Security Surface for Agent Guards

arXiv:2610.22244v1 Announce Type: cross Abstract: Step-level guards decide, before each tool call, whether an agent's next action is safe given the user's request and the trajectory so far. Users writ

Local Ai

LICOA: Learning Instruction-Conditioned Optical Actions for Mobile Spatial Augmented Reality

arXiv:2610.21282v1 Announce Type: cross Abstract: Spatial augmented reality (SAR) augments physical environments with projected content, but most systems assume that the projector pose, target region,

local-aiarxiv-cs-cv

Momentum, Commentary Tone, and Earnings Surprise in LLM Financial Forecasts

arXiv:2610.22043v1 Announce Type: cross Abstract: Which inputs drive a language model's financial forecast? We construct synthetic earnings announcements that independently reverse 20-day momentum, no

NeoHorse-1: Towards Recursive Self-Improvement via Agentic Post-Training with Routing Harness

arXiv:2609.08183v2 Announce Type: replace Abstract: Recursive self-improvement (RSI) requires a concrete mechanism through which an AI system observes its capabilities and converts that evidence into

safetyarxiv-cs-cl

No Single Tokenizer Feature Reliably Predicts Downstream Language Model Performance

arXiv:2610.21679v1 Announce Type: new Abstract: Tokenization is a critical component of the language modeling pipeline, but it is poorly understood. Tokenizers can impact downstream performance, but t

researcharxiv-cs-cl

OmicsBench: a long-horizon multi-turn benchmark of AI agents on research-study-scale bioinformatics

arXiv:2610.21074v1 Announce Type: new Abstract: Computational biologists increasingly delegate substantial analyses to AI agents in long-horizon, multi-turn sessions, yet most benchmarks of agents' bi

Perceptual Comparisons of Viscoelastic Renderings: Fractional- vs Integer-Order Models

model-releasesarxiv-cs-ro

arXiv:2610.21965v1 Announce Type: new Abstract: Realistic haptic rendering of viscoelastic materials requires reproducing time-dependent behaviors such as creep and stress relaxation, which arise from

Policy-Aligned Predictive Verification for Adaptive VLA Action-Chunk Execution

arXiv:2610.21434v1 Announce Type: cross Abstract: Vision-language-action (VLA) policies generate action sequences that reduce repeated inference, but deciding how much of each sequence to execute is c

safetyarxiv-cs-lg

PriorityFEC: Deadline-Aware Streaming FEC for KV-Cache Handoff in Edge Multi-Agent Systems

arXiv:2610.21390v1 Announce Type: cross Abstract: Multi-agent large language model (LLM) systems are increasingly deployed at the edge to support complex tasks that require collaboration and informati

Representation Redistribution in Sparse Autoencoders

arXiv:2610.21240v1 Announce Type: cross Abstract: Sparse autoencoders (SAEs) decompose language model activations into sparse features, but prompt transformations designed to preserve the underlying r

safetyarxiv-cs-ai

Retrieved but Not Read: Understanding Why Search-Augmented LLMs Fail to Use Available Evidence

arXiv:2610.21731v1 Announce Type: cross Abstract: Web search allows LLMs to access up-to-date information, but it remains unclear why they sometimes produce outdated answers even when relevant evidenc

Sign Betting for Misspecified Regression: Variance Adaptivity and Optimal Approximation

arXiv:2610.20903v1 Announce Type: new Abstract: We study bounded regression under model misspecification, evaluating a function f by its L_1(P_X) distance to the conditional mean eta(x)=E[Ymid X=x], r

applicationsarxiv-cs-lg

SWE-Nexus: Automated Composition of Software Engineering Tasks from Test Execution

arXiv:2610.20038v2 Announce Type: replace-cross Abstract: Historical code changes provide convenient task boundaries, but do not identify which production functions selected tests exercise or support

Tool-Augmented Large Language Models for Reliable Medical Calculation from Patient Notes

arXiv:2610.21477v1 Announce Type: new Abstract: Medical calculators translate raw patient observations into clinically actionable values and scores. However, applying these calculators directly to fre

VELD-QA: A Large-Scale Dataset for Visual Evidence Grounding and Question Answering in Long Documents

arXiv:2610.20964v1 Announce Type: cross Abstract: Existing long-document QA benchmarks rarely combine large-scale human-authored questions with detailed visual evidence annotations, limiting diagnosis

Agents

What Does History Add? Calendar Controls and Information-Timing Audits for University Trajectory Forecasting

arXiv:2610.21419v1 Announce Type: new Abstract: We audit the predictive value of historical summaries in a rule-based university trajectory simulator. Seventeen ridge specifications and persistence se

agentsarxiv-cs-lg

When Correctness Feedback Erases Evidence: Phase-Separated Completeness Auditing in Verus

arXiv:2610.20676v1 Announce Type: new Abstract: Correctness feedback suppresses the very evidence that specification-completeness auditing is designed to uncover: programs that satisfy a written contr

When do quantum views help? Complementary classical-quantum representations in multi-view fusion

arXiv:2610.21395v1 Announce Type: cross Abstract: Recent studies have reported promising results from quantum machine learning in drug discovery, healthcare, materials science and cybersecurity. Resea

When Models Misread Ink: Auditing VLMs for Arabic Calligraphy Recognition

model-releasesarxiv-cs-cv

arXiv:2610.22018v1 Announce Type: new Abstract: Arabic calligraphy is a high-stakes cultural visual domain for Muslim communities, yet VLM evaluations rarely test its fine-grained script traditions. W

Local Ai

Wolof-Adapted Industry-Ready Open-Weight Language Models

arXiv:2610.20824v1 Announce Type: cross Abstract: Wolof is spoken by 23M people, mostly in Senegal, but remains weakly supported by open-weight language models, while API models raise cost, governance

local-aiarxiv-cs-ai

WorldCast: Exploring Intelligent Visual Generation in World State Prediction

arXiv:2610.20731v1 Announce Type: cross Abstract: Humans interpret multimodal observations of the world and forecast how it will evolve. Recent multimodal generative systems have demonstrated similar

A Graph-Based Analysis of Comparability, Scale, and Attribution in Private Transformer Inference

arXiv:2610.19607v1 Announce Type: cross Abstract: Rising privacy concerns around large language models have accelerated research in private Transformer inference (PTI). Synthesizing the empirical clai

applicationsarxiv-cs-lg

A large-scale pipeline for LLM-assisted corpus annotation: variation and change in the English consider construction

arXiv:2510.12306v4 Announce Type: replace Abstract: As natural language corpora expand at an unprecedented rate, manual annotation remains a significant methodological bottleneck in corpus linguistic

applicationsarxiv-cs-cl

Accelerating the search for antimicrobial peptides with Google Antigravity

model-releasesgoogle-deepmind

Google Antigravity is an agentic science workbench that consolidates multiple biological databases into a single interface, enabling biologists to rapidly discover, evaluate, and prioritize therapeuti

AdaptFuse: Training-Free Sequential Preference Inference via Externalized Bayesian Computation

arXiv:2604.03925v2 Announce Type: replace Abstract: Large language models (LLMs) struggle to accumulate evidence across multiple rounds of user interactions, failing to update their beliefs in a manne

AegisVox: Factorized Privacy and Clinical Preservation for Long-Form Speech

arXiv:2610.14643v1 Announce Type: new Abstract: Clinical speech reveals identity through both spoken identifiers and voice, yet its pauses, delivery, and speaker roles can also carry useful clinical e

safetyarxiv-cs-ai

AgentStripes: An Evidence-Graded Maturity Model for AI Agents

arXiv:2610.18603v1 Announce Type: new Abstract: Benchmarks report what an Artificial Intelligence (AI) agent can do on a task suite, but not whether one released version has been engineered with the c

AM-Bench: Understanding Agents under Task Underspecification across Models and Harnesses

arXiv:2610.18090v1 Announce Type: new Abstract: Ambiguous or incomplete task specifications complicate both execution and clarification for language-model agents. Yet agents' responses to these inform

Annealed Levy Diffusion via Gaussian Surrogate Reverse Dynamics

arXiv:2610.19692v1 Announce Type: new Abstract: Gaussian diffusion models offer a practical framework for generative modeling. Denoising Levy Probabilistic Models (DLPM) introduce heavy-tailed perturb

researcharxiv-cs-lg

Are Multimodal Benchmarks Really Useful? Item-Level Multimodal Benchmark Diagnosis via Structure-Response Co-Calibration

arXiv:2610.14542v1 Announce Type: new Abstract: Multimodal benchmarks are central to evaluating whether models can reason over information from multiple modalities. However, existing literature has re

Are We Overthinking One-Step Generation?

arXiv:2610.20331v1 Announce Type: new Abstract: Diffusion and flow matching are now dominant frameworks for generative modeling in continuous domains, but they typically require multiple network evalu

researcharxiv-cs-lg

AtomSearch: Evidence-Grounded Tree Search for Agent Skill Optimization

arXiv:2610.19951v1 Announce Type: new Abstract: Agent skills adapt frozen models through editable external instructions. Revising several rules at once makes changes difficult to inspect, while judgin

Back-Translation Still Matters: Leveraging Monolingual Data for Low-Resource Machine Translation with LLMs

arXiv:2610.18818v2 Announce Type: cross Abstract: Pre-trained Large Language Model (LLMs) are effective on many tasks in zero-shot and few-shot settings. For machine translation of low-resource langua

BANGLAPLAN: A Benchmark for Constraint-Based Planning and Decision-Making in Bangla

arXiv:2610.19704v1 Announce Type: new Abstract: Large language models (LLMs) increasingly help users arrange appointments, plan journeys, and allocate limited resources. These tasks require models to

BenchPreS: A Benchmark for Context-Aware Personalized Preference Selectivity of Persistent-Memory LLMs

arXiv:2603.16557v2 Announce Type: replace-cross Abstract: Large language models (LLMs) increasingly store user preferences in persistent memory to support personalization across interactions. However,

Continuous Risk-Based Scheduling for Food Safety Inspection: Integrating Historical Decay with Event-Driven Consumer Review Signals

arXiv:2610.15191v1 Announce Type: new Abstract: Batch risk-based inspection scheduling struggles to detect a critical failure class: food facilities with recent clean inspections that subsequently dev

Convergence Rates of Gaussian Process Thompson Sampling for Efficient Global Optimization

model-releasesarxiv-cs-lg

arXiv:2610.19925v1 Announce Type: cross Abstract: Efficient global optimization aims to identify a global optimum of an expensive black-box function within a limited evaluation budget N. We investigat

Coupled Harness--Model Evolution for Scientific Equation Discovery

arXiv:2610.17740v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used to discover scientific equations from experimental data. Such a discovery system has two parts that m

safetyarxiv-cs-ai

Cross-Lingual TTS for Indian Languages: A Benchmark and Inference-Time Duration Estimation Strategies

arXiv:2610.14826v1 Announce Type: cross Abstract: Recent alignment-free Text-to-Speech (TTS) models achieve high synthesis quality by removing explicit alignment modules and simplifying training. Howe

Cross-Modal Latent Influence: Patch-Level Training Sample Attribution for Vision-Language Models

arXiv:2610.14190v1 Announce Type: new Abstract: Explaining a vision--language model (VLM) prediction requires tracing training-example influence to the multimodal evidence that mediates it, yet existi

D2: High-Capability Full-Duplex Spoken Language Models from Half-Duplex Backbones

arXiv:2610.17124v1 Announce Type: cross Abstract: Full-duplex spoken language models (FD-SLMs) support simultaneous listening and speaking, but often lag behind half-duplex spoken language models (HD-

Diagnosing and Correcting Reward Hacking in Verifiable Instruction Following

arXiv:2610.19525v1 Announce Type: cross Abstract: Verifiable instruction-following benchmarks emphasize output-constraint satisfaction but can overlook whether responses meaningfully answer the underl