AI archive
Every AI story, newest first. 91 in this archive.
Oct 12, 2026
MAGMA Bench Stress Tests Multi Robot Agents Under Live Failures
Researchers released MAGMA Bench on 9 October 2026 to score long horizon embodied agents when tasks fail, interrupt, or split across heterogeneous robots.
Oct 11, 2026
RobotWorld Shows Digital Agents Still Fail When Physics Must Compose
The RobotWorld arXiv benchmark reports that frontier agents can build perception pipelines yet still miss object state and completion checks across 84 embodied tasks.
Oct 10, 2026
NavGPT 3 Links Language Reasoning Threads to a Navigation Policy at Human Pace
NavGPT 3 pairs a hierarchical runtime with a 19 million example vision language action policy, reaching human level success on RxR CE while cutting reaction latency versus pure LLM control.
Oct 9, 2026
Memento 3 Stores World Models in Text Rulebooks Agents Can Rewrite
Researchers propose Memento 3, letting frozen LLM agents learn explicit world models through revisable natural language rulebooks and verified code compilation on ARC AGI 3 and Atari Pong.
Oct 8, 2026
Stateless Language Agents Keep Research State Outside the Chat Window
A new arXiv paper argues long horizon research agents fail when conversation history becomes the system of record, and proposes stateless workers with a harness owned search state.
Oct 6, 2026
VERA Builds 9000 Verifiable Agent Sandboxes That Co Evolve With Models
Researchers introduce VERA, a framework that scales resumable verifiable training environments and co evolves models with agent harnesses, reporting large gains on work and medical agent benchmarks.
Oct 5, 2026
SURGE Extracts Stronger Policies From Finished RL Runs Without Retraining
An arXiv paper posted 1 October 2026 shows SURGE can fuse checkpoints from completed reinforcement learning histories into policies that beat the best training checkpoint on math and coding benchmarks without additional training or extra inference tokens.
Oct 4, 2026
ActiveSaddler Co Evolves Agent Harness Curricula From Failure Patterns
Microsoft researchers posted ActiveSaddler on arXiv 2610.00906, framing harness optimization as a non stationary bandit over failure patterns. GAIA2 and Terminal Bench 2.0 Pass at 1 gains reached 4.4 and 7.5 points.
Oct 3, 2026
YouRA Tracks Hypotheses So Agent Papers Match Executed Evidence
A new arXiv paper introduces YouRA, a persistent state architecture that links manuscript claims to executed experiments and failure logs, improving end to end research agent reliability on MLR Bench.
Oct 2, 2026
VISTA Harness Turns Multimodal Models Into Long Horizon Visual Agents
MIT researchers show a visual memory harness lifts Claude Opus 5.0 to a perfect ARC AGI 3 score while cutting action counts, offering a practical path for enterprise GUI and game agents.
Oct 1, 2026
GUI Agents Need Action Based Memory Not Lookalike Pages
An arXiv paper from 30 September 2026 defines when two web pages count as the same state for GUI agent memory using action outcomes, not pixel similarity.
Sep 30, 2026
WEFT Paper Scales Tool Use Post Training Across Whole Agent Systems
An arXiv paper submitted 29 September 2026 argues agent reliability requires evolving the full tool stack, not just model weights, with measured gains on BFCL V4 and long horizon benches.
Sep 29, 2026
Fluctuation Supervision Cuts Causal Tabular Error Nearly 70 Percent
A September 2026 arXiv paper introduces fluctuation supervised pretraining for causal tabular models, reporting a 69.8 percent RMSE reduction versus latent supervision on large effect shifts.
Sep 27, 2026
ProCredit Paper Turns Acceptance Checks Into Turn Level Agent Rewards
Researchers show that rerunning task acceptance checks on intermediate agent states yields verified progress credit that beats outcome only reinforcement learning on AppWorld benchmarks.
Sep 26, 2026
GRAFT Paper Fixes Step Level Credit Assignment in Agentic RL
A 25 September arXiv paper introduces GRAFT, a graph based framework that assigns step level advantages in multi turn agent training without sampling every intermediate state.
Sep 25, 2026
Env Rethink Paper Shows Agent Failures Start in Noisy File Environments
Shanghai Jiao Tong University and Tencent Hunyuan researchers posted Env Rethink on 24 September 2026, reporting that noisy file environments cut agent pass rates from 83.9% to 57.6% across nine model configurations.
Sep 24, 2026
Agent Editing World Model Trades Observation Simulation for History Cleanup
A 24 September arXiv paper introduces AEWM and EditAct, which edit noisy agent histories instead of hallucinating tool outputs, lifting six benchmark averages by 3.2 to 6.7 points across three backbones.
Sep 23, 2026
Agensh Shows Agent Count Scales Coding Reliability Without a Central Orchestrator
Microsoft researchers report a decentralized multi agent harness scaling to 1,024 workers, raising ProgramBench pass rates by about 49% at 128 agents on GPT 5.6 Sol.
Sep 22, 2026
Proxifield Routes Multi Agent Teams by Semantic Proximity at Scale
An arXiv paper shows decentralized routing beats centralized star topologies as agent teams grow, retaining 73.6% reward under severe permanent failures.
Sep 21, 2026
Emergence World Shows Detection Without Containment in Multi Agent Runs
A 15 September arXiv paper reports 16 days of unsupervised multi agent worlds where frontier models detected threats yet still wrote adversarial content into persistent memory up to 46 hours later.