Search: “persistence”

Localized Adaptation Reveals Distinct Learning Signatures in Transformers

arXiv:2607.25663v1 Announce Type: new Abstract: Transformer adaptation is typically distributed across model depth, even when the intended change is narrow. We investigate how adaptation site shapes w

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Research

Measuring and Improving Behavioral Consistency in Large Language Models through Fact-Heuristic-Emotion State Enforcement

DGX agent

arXiv:2607.24765v1 Announce Type: cross Abstract: Large language models (LLMs) can give different answers to the same decision problem across runs, and reverse a decision when their own prior answer r

researcharxiv-cs-ai

29 Jul 2026

MemLens: A Value-Aware Memory Management System with Interactive Analytics for LLM-based Agents

arXiv:2607.25992v1 Announce Type: cross Abstract: Recently, memory management has become a key infrastructure for LLM-based agents, as it directly affects long-horizon reasoning, personalized response

Memory for Large Language Models

model-releasesarxiv-cs-cl

arXiv:2607.25380v1 Announce Type: new Abstract: Memory has evolved into a foundational architectural dimension in large language models (LLMs), shifting from an implicit byproduct of computation to a

SecDrift: Measuring Sector-Conditioned Security Drift in AI-Generated Code

model-releasesarxiv-cs-lg

arXiv:2607.25225v1 Announce Type: cross Abstract: LLMs are increasingly used for code generation in critical infrastructure, yet the security effect of domain-specific prompting is understudied. We pr

The openai agent sandbox escape actually has very real implications for the diffusion of AI in the enterprise. The incident showed the power…

The openai agent sandbox escape actually has very real implications for the diffusion of AI in the enterprise. The incident showed the power and capability of agents, and the need to harden systems an

agentsgary-marcus--x

Towards Understanding the Cognitive Habits of Large Reasoning Models

arXiv:2506.21571v3 Announce Type: replace-cross Abstract: Large Reasoning Models (LRMs), which autonomously produce a reasoning Chain of Thought (CoT) before producing final responses, offer a promisi

Applications

When Algorithms Meet Artists: Semantic Compression and Stake-holder Marginalisation in Public AI-Art Discourse (2013-2025)

arXiv:2508.03037v5 Announce Type: replace Abstract: Artists occupy a paradoxical position in generative AI. Their own work trains models that now compete with them, replicate their styles, and reshape

applicationsarxiv-cs-cl

A Few Words Go a Long Way: Language Guided Robot Policy Synthesis

arXiv:2607.23784v1 Announce Type: cross Abstract: While vision-language-action models have demonstrated impressive zero-shot manipulation capabilities, they remain fundamentally black box policies tha

A Frozen 12B Beats Frontier Models on Verified Work: 100% Accuracy, 0 Tokens, Bit-Exact, Forever

arXiv:2607.23806v1 Announce Type: cross Abstract: Improving a language model today means retraining it: enormous compute, a new opaque model each cycle, non-deterministic output. We take the opposite

A Mechanistic Perspective and Circuit-Guided Difficulty Metric for Unlearning

arXiv:2601.09624v2 Announce Type: replace-cross Abstract: Machine unlearning is becoming essential for building trustworthy and compliant language models. Yet unlearning success varies considerably ac

A Multi-stage Constrained Optimization Framework for Data-driven Problems

arXiv:2607.23480v1 Announce Type: new Abstract: Variational autoencoders (VAEs) transform high-dimensional, often noisy data into a compact latent representation, making downstream optimization more t

safetyarxiv-cs-lg

Applications

A scalable online machine learning approach for Stock Recommendation

arXiv:2607.23120v1 Announce Type: cross Abstract: Stock recommendation systems face the dual challenge of adapting to rapidly changing market conditions while maintaining low-latency predictions for e

applicationsarxiv-cs-ai

An Efficient and Effective Evaluator for Text2SQL Models on Unseen and Unlabeled Data

arXiv:2603.07841v2 Announce Type: replace Abstract: Recent advances in large language models have strengthened Text2SQL systems that translate natural language questions into database queries. A persi

researcharxiv-cs-cl

ARdena: Scenario-driven control of real-time LLM agents

arXiv:2607.22651v1 Announce Type: new Abstract: Large language models (LLMs) have enabled increasingly capable conversational agents, but reliably controlling their behavior in real-time interactive e

safetyarxiv-cs-ai

Bayesian Repetition Penalty: A Principled Adjacent-Conditional Framework for Reversing Attention Collapse in Autoregressive Language Models

arXiv:2607.22694v1 Announce Type: new Abstract: Attention collapse in autoregressive language models -- manifested as repetitive token loops where the model becomes trapped in self-reinforcing attract

Applications

BettiSplit: Topology-Guided Privacy-Aware Split Learning Against Feature Inversion and Gradient Leakage

arXiv:2607.24556v1 Announce Type: cross Abstract: Split learning enables collaborative model training by partitioning neural networks across clients and servers. However, improper split placement can

applicationsarxiv-cs-ai

Building AI That Works: ESnet's Pragmatic Approach to AI-Driven Operational Excellence

arXiv:2607.22948v1 Announce Type: cross Abstract: The ORBIT (Operations Responses and Business Intelligence Toolkit) project was initiated to assess agentic AI for the upcoming ESnet 7 initiative and

agentsarxiv-cs-ai

ConsistencyGate: Preventing Memory Contamination in LLM Agents via Self-Consistency Admission Control

arXiv:2607.22962v1 Announce Type: new Abstract: LLM agents that operate over many turns accumulate facts in an external memory store and reuse them as premises for downstream reasoning. A hallucinated

Local Ai

Distribution-Specific Curvature Control with Finite-Sample Guarantees for Open-Weight Safety

arXiv:2607.22929v1 Announce Type: new Abstract: A short fine-tuning run can undo the safety guards of an open-weight model---retraining a refusal-trained assistant to aid weapons development or produc

local-aiarxiv-cs-lg

Do Current Retrievers Cover All the Evidence? A Controlled Study of Conjunctive Cross-Page Retrieval

model-releasesarxiv-cs-cl

arXiv:2607.24165v1 Announce Type: cross Abstract: Finding a long document relevant to a multi-part request is not the same as establishing that it contains every requested piece of evidence. We study

Do LLM Debates Repeat Arguments Differently Across Languages?

arXiv:2607.23442v1 Announce Type: new Abstract: LLM debate is usually evaluated by final answers, but transcripts also reveal whether later turns develop new argumentative content or return to earlier

safetyarxiv-cs-cl

EditCLEVR: A Paired-Scene Intervention Benchmark for Compositional Faithfulness of Object-Centric Representations

model-releasesarxiv-cs-cv

arXiv:2607.22705v1 Announce Type: new Abstract: Object-centric learning aims to represent scenes as objects whose properties can be reused in new combinations. Existing evaluations usually score segme

Execution-Grounded Security Testing for Coding Agents in Software Engineering Pipelines

arXiv:2607.22569v1 Announce Type: new Abstract: Coding agents are increasingly integrated into system operations, where their tool use can directly modify project artifacts, execution environments, an

safetyarxiv-cs-ai

Flick: Few Labels Text Classification using K-Aware Intermediate Learning in Multi-Task Low-Resource Languages

arXiv:2506.10292v2 Announce Type: replace-cross Abstract: Training deep learning networks with minimal supervision has gained significant research attention due to its potential to reduce reliance on

From Execution to Capability: Scientific Experience Consolidation via Procedural Knowledge Synthesis

arXiv:2607.24459v1 Announce Type: new Abstract: Large language models increasingly solve scientific-computing tasks, but executable feedback from one problem rarely becomes durable capability on subse

Tutorials

Hiding in Plain Sight: An Effective Physical Adversarial Patch Attack against Visual-Infrared Fused Face Detection

arXiv:2607.23292v1 Announce Type: cross Abstract: Deep learning-based visual-infrared fused face detection models are increasingly deployed across a wide range of applications, yet they remain suscept

tutorialsarxiv-cs-cv

Kimi K3: Open Frontier Intelligence

model-releasesarxiv-cs-cl

arXiv:2607.24653v1 Announce Type: new Abstract: We introduce Kimi K3, a 2.8T parameter Mixture-of-Experts model with 104 billion activated parameters, native vision capabilities, and a 1-million-token

LEX-EC: A Lexical Evidence-Channel Audit Framework for Zero-Shot LLM Personality Classification in Black-Box Settings

arXiv:2607.24435v1 Announce Type: cross Abstract: Large language models may easily assign personality labels from text, but model interpretability remains an open problem. To address this gap, we intr

MobiWave: Dispatch-Oriented Graph Wavelets and Drift-Guided Selective Optimization for Autonomous Fleet Rebalancing

arXiv:2607.24365v1 Announce Type: new Abstract: Autonomous fleets enable mobility platforms to coordinate idle vehicles directly, making fleet-wide rebalancing possible. However, two obstacles limit r

safetyarxiv-cs-lg

New research from Meta and CMU. This one is on agentic context management for long horizon tasks. (bookmark it) Production agents accumulate…

New research from Meta and CMU. This one is on agentic context management for long horizon tasks. (bookmark it) Production agents accumulate context every turn. The usual fix compresses on a token thr

model-releasesdair-ai--x

Not All LLM Reasoning is Visible in the Chain-of-Thought

arXiv:2607.22925v1 Announce Type: cross Abstract: A key question for AI safety is whether a language model expresses all of its reasoning in its output tokens. We demonstrate a concrete failure mode w

Not Forgotten: Implementation and Evaluation of a Personalized Episodic Memory for the Humanoid Robot Head Kim

arXiv:2607.24190v1 Announce Type: cross Abstract: Social robots that rely on large language models for conversation are unable to retain information across sessions. This absence of memory violates so

OmniMate: Open-Ended Real-Time Streaming Audio-Visual Generation for Interactive Avatars

arXiv:2607.23023v1 Announce Type: new Abstract: Recent advances in diffusion-based generative models have enabled real-time audio-driven avatar generation and unified audio-visual synthesis, providing

researcharxiv-cs-cv

Outcome-Confounded Local Supervision in On-Policy Distillation

arXiv:2607.23731v1 Announce Type: new Abstract: On-policy distillation (OPD) trains a student on its own trajectories while a teacher supplies dense token-level likelihoods at student-visited prefixes

safetyarxiv-cs-lg

ParBench: A Benchmark for Reliable Evaluation of LLM Parallel Code Translation

arXiv:2607.22588v1 Announce Type: new Abstract: Modern compute-intensive software must migrate across a changing ecosystem of accelerators, programming APIs, compiler stacks, and portability layers, i

Progress-conditioned Group Policy Optimization for Long-Horizon Agentic Tasks

arXiv:2607.22724v1 Announce Type: cross Abstract: Group-based policy optimization has been increasingly used to train large language model (LLM) agents from sparse outcome rewards by comparing traject

safetyarxiv-cs-ai

Hardware

Rendering on Real Silicon: GPU Render-Timing as a Passive, AI-Resistant CAPTCHA Signal

arXiv:2607.23389v1 Announce Type: cross Abstract: Conventional CAPTCHAs pose puzzles that modern AI systems increasingly solve, while behavioral and cryptographic-attestation defenses carry privacy or

hardwarearxiv-cs-lg

Risk reversal for least squares estimators under nested convex constraints

model-releasesarxiv-cs-lg

arXiv:2601.16041v2 Announce Type: replace-cross Abstract: In constrained stochastic optimization, one expects that restricting the feasible set, provided it still contains the true parameter, should n

Scientific Computing in the Age of Agentic AI

Human

OpenAI field report: eight life-sciences projects used Codex/Claude Code agents; bottleneck shifts from coding to verification.

agentssciencecodex

SpecBox: Speculative Sandbox Scheduling for Efficient LLM Agent Serving

arXiv:2607.23933v1 Announce Type: cross Abstract: As LLM agents increasingly rely on the Model Context Protocol (MCP) to invoke isolated external sandboxes, disaggregated sandbox deployment introduces

agentsarxiv-cs-ai

Stability of AI Governance Systems: A Coupled Dynamics Model of Public Trust and Social Disruptions

arXiv:2603.20248v2 Announce Type: replace-cross Abstract: AI systems are increasingly entrenched in public governance, yet scholarship lacks formal tools to determine when deviations of public trust i

Steerable Chatbots: Exploring Personalization Control Interfaces via LLM Activation Steering

arXiv:2505.04260v3 Announce Type: replace-cross Abstract: Personalizing LLM responses typically requires users to articulate their preferences through prompting, which can be burdensome at cold start

Stress-Testing EEG Foundation Models for Clinical Decoding: Dataset Identity and Targeted Negative Controls

arXiv:2607.24519v1 Announce Type: cross Abstract: Pretrained EEG foundation models are increasingly proposed for clinical decoding, but their transfer across populations and robustness to negative con

Success Is Not Self-Explanatory: Auditing Success Provenance in Agent Evaluation

arXiv:2607.24054v1 Announce Type: new Abstract: A correct answer can conceal why an agent succeeded. Once agents change their information state during evaluation, correctness no longer distinguishes i

Topological Data Analysis and Graph-Theoretic Approaches for Tennis Match Prediction

arXiv:2607.23509v1 Announce Type: new Abstract: We present two approaches for predicting tennis match outcomes using topological data analysis and graph theory on ATP singles matches from 2000-2025. T

researcharxiv-cs-lg

Variational-Ising-Attention (VIA):TailoredAttentionMattersfor Science

arXiv:2607.23634v1 Announce Type: cross Abstract: Attention enables context modeling via query-key scoring with softmax normalization. Driven by industrial long-context demands, mainstream research ha

Tutorials

When Activation Oracles Learn Not to Read: Concept-Specific Blind Spots in Fine-Tuned Oracles

arXiv:2607.23379v1 Announce Type: cross Abstract: Activation Oracles (AOs) are language models trained to answer natural-language questions about another model's internal activations. They offer a fle

tutorialsarxiv-cs-ai