Search: “Lean”

Smart search ranks by meaning as well as keywords (one row per story, last 45 days).

How well do agents use test/verification techniques?

Dan Luu's eval finds coding-agent testing instructions (TDD, formal methods, PBT, skills) mostly fail to beat defaults on Zstd implementation correctness.

The author ran 26 prompt conditions plus 4 skills on a Zstd-in-Rust implementation eval using codex with GPT-5.6, testing TDD, fuzzing, property-based testing, formal methods (Lean 4, TLA+, Verus, Kani, SMT solvers) and community skills. Nothing dramatically outperformed the default no-instruction condition, which did above average; at xhigh effort, fuzzing and PBT conditions did slightly better than formal methods. Pre-registered predictions included TDD underperforming and popular test skills (ECC, Hegel, Trail of Bits) not outperforming. Results are averages of 80 runs per condition plotted against cost.

Hacker News · security·Sep 8, 02:58 UTC · 8d agoAI researchHN 131↑ · 50 comments

38

StochBench: A Domain-Specific Benchmark for Stochastic Processes in Lean

StochBench introduces 450 graduate-level stochastic-processes problems in Lean 4; an Opus 4.8-based agent proved 34.9% under a 15-minute limit.

StochBench is a Lean 4 benchmark of 450 graduate stochastic-processes problems, each paired with its natural-language source, covering Markov chains, renewal processes, martingales, Brownian motion, stochastic calculus, weak convergence, and Poisson and continuous-time Markov processes. The benchmark addresses field-specific applied mathematics underrepresented in Mathlib, unlike competition-math-dominated suites such as IMO and Putnam collections. An Opus 4.8-based agent achieved a 34.9% proof rate (157/450) under a 15-minute per-problem limit, showing the benchmark remains challenging for advanced provers.

Hugging Face daily papers·Sep 7, 20:00 UTC · 8d agoAI research

25

The 12 Best Endpoint Detection & Response (EDR) Solutions, Compared and Priced

An editorial scorecard ranks 12 EDR platforms, with CrowdStrike and SentinelOne tied at 8.6/10 and telemetry retention identified as the hidden cost driver.

An editorial comparison scores twelve EDR platforms on detection, response, analyst burden, pricing transparency, and coverage. CrowdStrike and SentinelOne tie at 8.6/10, with Microsoft Defender for Endpoint close behind at 8.5 and described as effectively free in Microsoft 365 E5 estates. The guide argues that telemetry retention, not per-endpoint price, drives real cost, with fully-priced quotes frequently diverging 2-3x from headline rates. Managed detection offerings, including Cynet's bundled 24/7 SOC, factor into the buyer-fit rankings.

GBHackers·Sep 9, 07:46 UTC · 7d agoIndustry 2 sources

15

Decomposition Buys Integrity, Not Yield

Study of 600 production deep-research traces finds agent-tree decomposition loses findings at rate N^(1-δ); flat architectures maximize yield.

The paper models multi-agent decomposition as a tree where an agent holding b items retains each with probability r(b); with r(b)=1/b every tree delivers exactly one finding regardless of shape. Analysis of 600 production deep-research traces estimates delta=0.34 retention decay, and 1,012 annotated traces show one brief in sixteen goes off-target per tier, giving an alignment penalty of 0.536. Depth still cuts root context exposure from N to N^(1/k) and is cheaper at scale, with a hazard model over 743,819 production tool calls showing delegation is an opening move rather than a response to filling context.

arXiv cs.AI / cs.LG / cs.CL·Sep 15, 17:04 UTC · 1d agoAI research

30

[AINews] GPT-6 Astra: OpenAI’s biggest LLM launch of all time

OpenAI launched GPT-6 Astra, its new flagship model, claiming state-of-the-art computer use, software engineering, math, and cybersecurity capabilities.

OpenAI launched GPT-6 Astra as its new flagship model, describing it as its most intelligent and aligned model with state-of-the-art computer use, software engineering, and math/science capabilities. Pricing is $10/$50 per 1M input/output tokens standard ($20/$100 fast tier), rolling out first to limited organizations, then ChatGPT Plus/Pro/Business/Enterprise, the API, and AWS. OpenAI claims 99.9% on ARC-AGI-3, 98% on FrontierMath Tier 4, and 100% on ExploitBench. Artificial Analysis scored Astra 67 on the Coding Agent Index and 61 on the Intelligence Index, behind Claude Fable 5.1, and the system card drew attention for reporting decreased chain-of-thought monitorability despite alignment gains.

Latent Space·Sep 4, 05:18 UTC · 12d agoModel release3

94

IdeaAMBIG: Benchmarking Implementation-Critical Gaps in Research-Idea Specifications

IdeaAMBIG benchmark with 660 instances measures whether LLMs can spot and fix underspecified research-method details for faithful implementation.

Researchers introduce IdeaAMBIG, a benchmark of 660 evidence-grounded instances (163 real-world gaps from reproducibility reports and GitHub issues, 497 controlled synthetic gaps) built from papers, codebases, and reproduction artifacts. It evaluates codification-readiness assessment, defect localization, and clarification action generation. Across 13 LLMs, the best model achieved only 9.6% Macro Defect Recovery Rate on real-world instances but 80.6% clarification success when given the annotated defect. An oracle study showed gold resolutions raise the codification-ready rate from 14% to 98%, identifying defect localization as the main bottleneck.

arXiv cs.AI / cs.LG / cs.CL·Sep 9, 17:59 UTC · 6d agoAI research2

28

Researchers open-source a Wi-Fi cyber range for security training

NTNU and Aegean researchers open-source a software-emulated Wi-Fi cyber range using mac80211_hwsim with LLM-assisted scenario building.

Researchers from the Norwegian University of Science and Technology and the University of the Aegean published a design and prototype for a cyber range dedicated to IEEE 802.11 security training, emulating access points and clients with mac80211_hwsim, Linux namespaces, hostapd, wpa_supplicant, dnsmasq, and FreeRADIUS. The platform bundles Aircrack-ng, Wireshark, and custom tools WPAxFuzz and Bl0ck, and can convert plain-language scenario descriptions into deployable definitions via a locally hosted Llama model. A working prototype covering scenario creation and deployment is on GitHub; monitoring, access control, and orchestration zones remain future work.

Help Net Security·Aug 24, 04:21 UTC · 23d agoTools1

28

IdeaAMBIG: Benchmarking Implementation-Critical Gaps in Research-Idea Specifications

IdeaAMBIG benchmark of 660 specification-gap instances shows LLMs localize implementation-critical research gaps poorly, with best model at 9.6% defect recovery.

IdeaAMBIG is a benchmark of 660 evidence-grounded instances evaluating whether research-method specifications provide enough information for faithful implementation: 163 real-world gaps from reproducibility reports and GitHub issues plus 497 controlled synthetic gaps. It tests codification-readiness assessment, defect localization, and clarification action generation across 13 LLMs. The best model achieves only a 9.6% Macro Defect Recovery Rate on real-world instances, though 80.6% clarification success when given the annotated defect, and an oracle study shows gold resolutions raise codification-ready rates from 14% to 98%. Defect localization emerges as the main bottleneck across all evaluated models.

Hugging Face daily papers·Sep 8, 20:00 UTC · 7d agoAI research

30

Tell HN: OpenAI brings back 5 hour limit for plus and business standard users

OpenAI reinstated a 5-hour usage limit for Plus and Business Standard subscribers, sparking Hacker News debate about subsidized inference pricing and subscription value.

A Hacker News discussion reports that OpenAI has brought back a 5-hour usage limit for Plus and Business Standard users. Commenters debated whether cheap tokens are a subsidized customer acquisition strategy, whether AI companies have sustainable margins on inference, and how switching costs between providers affect dependency. The thread reflects community sentiment on pricing and usage caps rather than an official policy announcement.

Hacker News · security·Sep 7, 16:40 UTC · 9d agoAI industryHN 126↑ · 142 comments

35

Cadence: Error-Bounded Lossy Compression of Demand Time Series with a Time-Series Foundation Model

Cadence pairs Google's 330M-parameter TimesFM-3 foundation model with adaptive arithmetic coding, gaining 13-28% on 2026 demand series over classical predictors.

Cadence is an error-bounded lossy compressor for numeric time series combining the 330M-parameter Google TimesFM-3 foundation model with an adaptive arithmetic coder, guaranteeing a per-sample error bound. On 49 EIA-930 balancing-authority demand series from 2026 it gains 13.3% over the best of six classical predictors and 28.3% on 50 MTA ridership series, winning all 297 series-tolerance pairs with a 21.4% median gain. The paper also reports negative results, including that foundation models add negligible value for lossless coding and that PyTorch predictions are not bit-identical across batch sizes.

Hugging Face daily papers·Sep 4, 20:00 UTC · 11d agoAI research1

25

Top 10 Best Patch Management Software in 2026

Roundup ranks 2026 patch management software, favoring Automox, Action1's free tier and Tanium, and warns buyers to vet patching platform security.

This buyer's guide ranks ten patch management tools for 2026, placing Automox first for cloud-native patching, Action1 for a genuinely free small-estate tier and Tanium for patching hundreds of thousands of endpoints. It contextualizes the category with the 2021 Kaseya VSA ransomware supply-chain incident and the 2020 SolarWinds Orion compromise, arguing the security of the patching platform itself must be part of evaluation. It also notes Ivanti products have repeatedly appeared in CISA's Known Exploited Vulnerabilities catalog.

Cyber Security News·Sep 9, 07:54 UTC · 7d agoIndustry

14

K-Bench: a clinically calibrated benchmark for evaluating large language models in high-risk mental health conversations

Clinician-calibrated K-Bench evaluates 125 LLM configurations on 200 high-risk mental health vignettes, exposing wide variation in suicide and violence risk handling.

K-Bench is a clinician-calibrated, protected benchmark evaluating 125 model configurations from 33 base models across 14 providers on 200 multi-turn vignettes covering suicide, self-harm, domestic violence, substance misuse and no-risk presentations. A frozen GPT-4o judge achieved 94.2% exact agreement with clinician consensus across 6,751 eligible comparisons from 151 clinician-rated transcripts. Leading models combined supportive conversation with combined-risk scores above 95, while risk exploration varied substantially among weaker configurations; therapeutic prompting helped weaker models and elevated reasoning produced no average improvement. A continuously updated public leaderboard is hosted at k-bench.ai with protected test materials.

arXiv cs.AI / cs.LG / cs.CL·Sep 14, 16:49 UTC · 2d agoAI safety & security

46

When the Whole Company Adopts AI: What It Does to Your SOC

Analysis of 16.9 million SOC alerts finds AI-related alerts at 0.43%, growing 685% since February, with 94.1% noise and 0.02% real attacks.

A review of roughly 16.9 million SOC alerts found about 73,000 (0.43%) were AI-related, a share that grew 685% between February and June 2026. Of AI-related alerts, 94.1% were noise, 5.8% genuine risks, and 0.02% real attacks; 79.8% received benign verdicts, 81.7% were automatically suppressed, and only 5.4% reached a human analyst. The only confirmed attacks were phishing campaigns that weaponized AI brand names as lures, while developer coding agents spawning shells and reading credential stores routinely tripped detections written before AI agents existed.

The Hacker News·Sep 12, 10:24 UTC · 4d agoResearch1

45

Embodied-BenchForge: A Closed-Loop Agentic Workflow for Embodied Benchmark Construction

Embodied-BenchForge automates embodied benchmark construction via closed-loop synthesis with verification and repair, yielding seven benchmarks for MLLM evaluation.

Embodied-BenchForge is an agentic framework that transforms user-specified evaluation intents into complete embodied benchmark artifacts via Closed-Loop Benchmark Synthesis. Skill-Orchestrated Artifact Synthesis composes typed reusable skills while an artifact dependency graph records intermediate outputs; Requirement-Guided Verification and Repair triggers local re-execution or upstream rollback on failures. It constructs six Offline EQA benchmarks plus one interactive benchmark with 220 executable tasks, distinguishing MLLM and embodied agent capabilities in observation-based understanding and closed-loop execution.

arXiv cs.AI / cs.LG / cs.CL·Sep 11, 17:16 UTC · 4d agoAI research

18

AD Rights Management Service (Part 2): Extraction, Offline Decryption, and the Unrotatable Key

Huntress research shows AD RMS SLC root key is unrotatable and never expires, so its compromise permanently exposes all RMS-protected documents.

Part 2 of Huntress's AD RMS series details server-side attacks: extracting the Server Licensor Certificate (SLC) private key and performing offline decryption of protected documents. The SLC key has no expiry or rotation mechanism, with a 255-year certificate validity (2002–2258), so whoever recovers it can decrypt every document the deployment ever protected, indefinitely. The author released SharpRMS, a unified tool combining the 2016 DisARMS client-side attacks with new server-side key extraction and decryption capabilities. The research frames the SLC as comparable to KRBTGT and the DPAPI domain backup key, though not equivalent to domain compromise.

Huntress·Sep 10, 04:00 UTC · 6d agoResearch

58

FlowBalance: Verifier-Grounded Self-Improvement from On-Policy Reasoning Experience

FlowBalance is a verifier-grounded self-improvement method that beats FlowRL on Qwen3-4B and Qwen3-8B math reasoning while improving training stability.

FlowBalance calibrates dense self-guidance scores with verifier-derived group advantages: guidance is retained on positive-advantage trajectories, reversed on negative-advantage trajectories, and disabled when rollout groups show no outcome preference. The method exponentially reweights a reference policy via trajectory balance, with guarantees including within-group contrast preservation and a minimum-change reverse-KL characterization. On mathematical reasoning it outperforms FlowRL on Qwen3-4B and Qwen3-8B, trains faster and more stably, avoids direct OPSD's response-length collapse, and shows higher correct-strategy diversity on AIME24.

Hugging Face daily papers·Sep 2, 20:00 UTC · 13d agoAI research

32

Kids’ online safety bill faces dim prospects of passage this session despite progress

Kids Online Safety Act clears Senate committee but passage looks unlikely this session amid House-Senate deadlock over the duty-of-care provision.

KOSA advanced out of the Senate Commerce Committee, but the chambers remain split on a duty-of-care provision requiring platforms to act with reasonable caution to prevent foreseeable harm, which House leadership opposes over First Amendment and negligence-lawsuit concerns. The Senate passed KOSA 91-3 last Congress before it died in the House, and the House passed its own version without the duty of care in June as part of a larger package, after stripping a state-law preemption provision. Observers say even Senate passage this year is a struggle given the short calendar, with a lame-duck window between November and January the more plausible path, and Majority Leader John Thune controlling whether a roll-call vote happens.

The Record·Aug 24, 00:00 UTC · 23d agoPolicy & legal

45

Evaluating Time-Series Foundation Models and Multimodal Dietary Context for CGM Forecasting

Study finds zero-shot time-series foundation models underperform on CGM forecasting; fine-tuned Chronos-Bolt cuts RMSE up to 18.4% and dietary context adds signal.

The paper evaluates time-series foundation models for continuous glucose monitoring forecasting across eight public datasets covering Type 1 diabetes, Type 2 diabetes, and non-diabetes populations. Under a unified protocol, zero-shot foundation models did not consistently outperform baselines like Elastic Net and PatchTST, but lightweight fine-tuning did, with fine-tuned Chronos-Bolt reducing RMSE by 6.5%-18.4% in the T1D cohort and 8.6%-18.2% in the non-diabetes/T2D cohort. A residual-based fusion framework adding dietary context from CGMacros reduced overall RMSE by about 3% and postprandial RMSE by about 15% versus CGM-only baselines.

arXiv cs.AI / cs.LG / cs.CL·Sep 10, 17:43 UTC · 5d agoAI research

28

Human Attacker Hits Machine-Speed Exploitation of Marimo RCE

Sysdig details a human attacker chaining pre-auth RCE in Marimo (CVE-2026-39987) to AWS credentials and a bastion host in eight seconds.

Sysdig's Threat Research Team documented an intrusion where an operator exploited CVE-2026-39987, a pre-auth RCE in the Marimo notebook terminal WebSocket endpoint, gaining an interactive shell with no credentials. The attacker harvested AWS credentials from the process environment and Redis backend, retrieved an SSH private key from AWS Secrets Manager, and authenticated to an internet-reachable bastion host. The chain fired in eight seconds after roughly four hours of toolkit building, logging over 850 commands during a nine-hour session. The flaw, fixed in Marimo 0.23.0, has been on CISA's KEV catalog with a May 7, 2026 federal remediation deadline.

Infosecurity Magazine·updated Sep 15, 11:52 UTC · 1d ago·first Sep 14, 14:15 UTC · 2d agoExploit / PoC in the wild 4 sourcesCVE-2026-39987

60

Context Engineering Inside the Harness: 4 Mechanisms That Beat Context Overflow and Goal Loss on Long-Horizon Tasks

Survey of four harness mechanisms—context budgeting, compaction, todo-state, and memory—that keep long-horizon LLM agents on task across 200+ tool calls.

The article details how agent harnesses, not larger context windows, solve context overflow and goal loss on long-horizon tasks, citing Chroma's Context Rot report showing 18 LLMs (GPT-4.1, Claude 4, Gemini 2.5, Qwen3) degrade on long inputs. Concrete implementations include LangChain Deep Agents offloading tool responses over 20,000 tokens to the filesystem and truncating old tool calls at 85% window usage, and Claude Code capping auto memory at 25KB while re-reading the 5 most recently modified files after compaction. OpenAI's Responses API now offers server-side compaction via context_management with a standalone /responses/compact endpoint, which Codex uses for long-running coding tasks. Manus reports a roughly 100:1 input-to-output token ratio per ~50-tool-call task, motivating todo.md state recitation to prevent goal drift.

MarkTechPost·Sep 13, 05:56 UTC · 3d agoAI research1

52

One Attacker Has Scraped Both Salesforce and ServiceNow Portals Since 2025

Reco tracks the City Forum campaign: one Contabo VPS has scraped Salesforce and ServiceNow guest portals across finance, telecom and public sector since March 2025.

Reco researchers identified infrastructure at 158.220.87.79, a Contabo VPS using Go's default net/http user agent, that has harvested records from Salesforce Experience Cloud and ServiceNow customer portals across telecom, financial services, enterprise software and public sector targets for over a year. The tool abuses over-privileged persistent guest users, enumerating Salesforce Aura sites (560,000+ events at one target), probing Lightning Web Runtime via the UI-API across API versions v56.0-v66.0, and hitting the poorly documented ServiceNow Service Portal search endpoint. No named victims or actor attribution yet, and activity is ongoing with rising volume. Defenders are advised to hunt the Go-http-client user agent and specific IP in Salesforce Event Monitoring logs and syslog_transaction on ServiceNow, then tighten guest sharing rules and Knowledge Base read criteria.

The Hacker News·Aug 18, 11:30 UTC · 29d agoThreat actor in the wild1

68

ActReview: Rebuttal-Guided Training Data and Rubric Rewards for Actionable Peer Review Generation

ActReview post-trains Qwen3-8B-Base on OpenReview rebuttals to generate actionable peer-review feedback with grounded revision suggestions, benchmarked on 1,000 curated instances.

The paper defines Actionable Peer-review Generation as diagnostic claim generation plus revision suggestion generation and introduces ActReview, a rebuttal-guided post-training framework. From OpenReview review-rebuttal threads the authors build ActReview-40K, aligning reviewer weaknesses with author responses grounded in localized paper evidence, and post-train Qwen3-8B-Base with multi-task SFT followed by GRPO using weakness-specific rubric rewards. They also release ActReview-Bench, a human-curated 1,000-instance benchmark, on which ActReview outperforms prior specialized review-generation models on actionability and grounding while remaining competitive with strong prompt-based LLMs. Human evaluation confirms improved revision usefulness but identifies a remaining gap in technical accuracy.

Hugging Face daily papers·Sep 7, 20:00 UTC · 8d agoAI research

30

ToxicPanda 2.0 Gets a Major Upgrade, Expanding Attacks Across 16 Countries

Zimperium documents ToxicPanda 2.0, an Android banking trojan now targeting 349 financial institutions in 16 countries via ADB privilege escalation and overlay credential theft.

Zimperium's zLabs documented ToxicPanda 2.0, an Android banking trojan expanding from 16 targeted apps to 349 financial institutions across 16 countries, with 167 remote commands. It poses as a dropper, abuses VPN permissions to block Google Play Protect while installing a hidden payload, then uses the Accessibility Service for screen monitoring and overlay-based credential theft. It automates enabling Android Wireless Debugging and completes the pairing handshake to gain ADB shell access for privilege escalation, and overlays fake lock screens to steal device PINs. Previously unfinished commands are now operational and samples are served from AWS-hosted storage buckets.

Security Affairs·Aug 22, 16:50 UTC · 25d agoMalware in the wild

65

StudyBench: Can Self-Evolution Squeeze Textbooks for Olympiad Capability?

Researchers introduce StudyBench, a physics benchmark showing self-evolution gains on textbook problems rarely transfer to olympiad-level questions.

StudyBench is a controlled physics benchmark splitting test data into an Application Set of difficult textbook problems and a Transfer Set of olympiad-level problems. Across three base models, representative self-evolution methods improved on the Application Set but rarely transferred to the harder Transfer Set. A guidance ablation reveals a Guidance Gap, and every method hits a Compute Plateau, indicating the remaining limits are method problems rather than data or compute problems.

Hugging Face daily papers·Aug 31, 20:00 UTC · 15d agoAI research

28

OpenAI’s Sam Altman says it would be ‘ill-advised’ to go public in 2026

Sam Altman says OpenAI will not IPO in 2026, calling it 'ill-advised' given safety concerns and market conditions.

OpenAI CEO Sam Altman stated the company will not go public in 2026 despite having confidentially filed for an IPO. He cited AI safety considerations and societal readiness, saying 'right now would be an ill-advised moment to go public.' The New York Times previously reported OpenAI had hired bankers and lawyers targeting a Q3/Q4 2026 listing but was leaning toward 2027 due to tech stock volatility and financial challenges.

TechCrunch · AI·updated Sep 12, 21:16 UTC · 3d ago·first Sep 12, 20:19 UTC · 3d agoAI industry 2 sources

38

Competence-Gated Pooling of Language Models and Priors for Event Forecasting

Paper proposes a competence gate pooling language model forecasts with external priors, improving Brier score from 0.0771 to 0.0732 across 2,357 binary questions.

The paper defines a language model's relative competence as its marginal value beyond an available external forecast, and derives conditions under Brier loss where model disagreement improves that forecast. A competence gate estimates domain-level source weights from resolved outcomes, shrinks uncertain estimates toward a global weight, and recalibrates the pooled forecast. Across 2,357 resolved binary questions and five language models, it improves the external baseline from 0.0771 to 0.0732 Brier and beats global forecast combinations, though it defers to the market on ForecastBench. Across four Qwen models, verbal confidence failed to identify when the model outperformed the external forecast, while outcome-estimated competence supported better abstention.

Hugging Face daily papers·Sep 9, 20:00 UTC · 6d agoAI research

38

Google researchers uncover criminal zero-day exploit likely built with AI

Google links a likely LLM-built criminal zero-day for an open-source admin tool to planned mass exploitation and maps AI-assisted threats.

Google Threat Intelligence Group linked a zero-day exploit for a popular open-source web-based administration tool, enabling 2FA bypass with valid credentials via a semantic logic error, to a criminal group, citing educational docstrings, a hallucinated CVSS score, and textbook Python as signs of LLM authorship; the vendor was notified before a planned mass exploitation campaign. The report also details Russia-nexus malware families CANFAIL and LONGSTREAM using AI-generated decoy code, the PROMPTSPY Android backdoor driving the UI through the Gemini API, APT27 using Gemini to build relay tooling, and the TeamPCP (UNC6780) supply chain compromise of LiteLLM and Trivy repositories that planted the SANDCLOCK credential stealer.

Help Net Security·Aug 24, 04:21 UTC · 23d agoThreat actor

78

ToolLoop: Closed-Loop Tool-Use Data Synthesis via Decomposed Generation and Dynamic Self-Feedback

ToolLoop introduces a closed-loop synthetic data framework whose 11K examples lift a 4B model to 86.40% on BFCL tool-use evaluation.

ToolLoop decomposes tool-use data synthesis into function-name sampling, backward derivation of user queries, and forward derivation of tool calls, with dynamic self-feedback at each stage. This shifts the paradigm from generate-then-filter to generate-verify-refine, reducing inefficient and imbalanced synthetic data. A 4B model trained on 11K synthetic examples reaches 86.40% accuracy on BFCL non-reasoning mode (86.07% in an Isolate variant excluding BFCL-overlapping functions) and 72.1% on ACEBench using only 18.3% of baseline training data.

arXiv cs.AI / cs.LG / cs.CL·Sep 8, 17:26 UTC · 7d agoAI research1

30

🔬“We have foundation models for language, not for physics” — Anima Anandkumar, Bren Professor of Computing

Caltech professor Anima Anandkumar discusses Neural Operators and FourCastNet for physics modeling, arguing inductive biases beat pure token scaling.

Anima Anandkumar, Bren Professor at Caltech and co-founder of Accelerated Understanding, describes Fourier Neural Operators that learn in frequency and spherical-harmonic domains to model weather, fusion, and fluid or heat flow. Her team built FourCastNet 3, a global weather model competitive with physics-based simulations that runs on consumer-grade GPUs. She also introduced TorchLean, a framework for writing PyTorch-style networks inside the Lean proof assistant for formal verification, and was appointed to the United Nations Scientific Advisory Board. She argues physical domains resist scaling due to tiny datasets and context lengths in the hundreds of billions, so progress comes from built-in structure and physical priors.

Latent Space·Aug 26, 15:15 UTC · 21d agoAI research1

42

Type Diversity Enables Transformers to Generalise Compositionally

Researchers show lexical-versus-structural compositional generalization gaps in Transformers stem from type diversity imbalance in datasets, not architectural limits.

The paper argues that Transformers' difficulty with structural compositional generalization is an artifact of low structural type diversity in prior benchmark datasets rather than an architectural limitation. Using Grammatical Framework, the authors create linguistically diverse variants of COGS and SLOG. They find type diversity correlates with compositional generalization equally in lexical and structural test cases, contradicting previous claims that compound divergence explains task difficulty.

arXiv cs.AI / cs.LG / cs.CL·Sep 11, 17:59 UTC · 4d agoAI research

20

Controversy over OpenAI's Maths Breakthrough

OpenAI claims its internal model proved the Navier-Stokes equations 'blow up' — a Millennium Prize Problem — amid allegations it borrowed mathematicians' methods.

OpenAI announced that an internal model produced a proof, certified in the Lean proof assistant, showing the Navier-Stokes equations can 'blow up,' implying infinite fluid speeds — a claimed solution to one of the seven $1-million Millennium Prize Problems. Mathematician Tristan Buckmaster alleged OpenAI, after learning of progress by him and Anthropic employee Levent Alpöge on 'blowing up' the related Euler equations, adopted a similar 'forcing' method; OpenAI's Sébastien Bubeck denied this, saying the model independently solved Euler by different means and produced the full Navier-Stokes proof over one weekend. Mathematicians including Diego Córdoba, co-developer of the forcing approach, remain cautious, and the community is still evaluating the competing proofs.

Hacker News · security·Sep 8, 16:57 UTC · 8d agoAI researchHN 21↑ · 5 comments1

62

TART: A Modular Tool for Technique-Aware Audio-to-Tablature Guitar Transcription

TART pipeline transcribes guitar audio to tablature with fingering and technique labels, outperforming prior baselines by up to 8.5 Tab F1 points.

TART is a modular four-stage pipeline for audio-to-tablature guitar transcription, combining an audio-to-MIDI transcription model, an expressive technique classifier, an audio-conditioned T5 encoder-decoder for string-fret assignment, and an automated tablature generator. In zero-shot evaluation on GuitarSet, EGDB, and two noisy augmented benchmarks, it averages 81.35% audio-to-MIDI F50 (+6.67 over the best baseline), 71.8% string-fret Tab F1 (+8.5), and 54.08% end-to-end Tab F1. The authors state it is the first framework to generate tablature with both fingering and expressive technique annotations directly from guitar audio.

arXiv cs.AI / cs.LG / cs.CL·Sep 10, 17:55 UTC · 5d agoAI research2

26

GTIG AI Threat Tracker: From Prompting to Autonomy – The Evolution of Adversarial AI

GTIG's Q2 2026 tracker shows adversaries adopting agentic AI workflows, including credential harvesting in under six hours and supply chain attacks by UNC6780.

Google Threat Intelligence Group's Q2 2026 report documents adversaries moving from basic prompting to agentic AI workflows and automation, including a cloud compromise followed by agent-enabled mass credential harvesting executed in under six hours. It tracks financially motivated actor UNC6780 (TeamPCP) conducting large-scale open source supply chain compromises across PyPI, npm, and Docker Hub since March 2026, deploying credential stealers. The report also highlights growing targeting of proprietary AI models, source code, prompts, and API credentials, plus LLMJacking practices where adversaries steal developer credentials or hijack cloud infrastructure to run unauthorized AI workloads.

Google Threat Intelligence·Sep 8, 14:00 UTC · 8d agoThreat actor in the wild1

65

General Quantification of Covariate and Concept Shifts

Paper proposes γ*-concept shifts via entropic optimal transport, deriving estimable generalization bounds unifying covariate and concept shift under distribution shift.

The authors show existing definitions of concept shift break when source and target supports mismatch and propose γ*-concept shifts grounded in entropic optimal transport. They derive a general error bound covering broad loss functions, label spaces and stochastic labeling, plus estimators with concentration guarantees. The resulting DataShifts algorithm quantifies distribution shifts and estimates the error bound in most applications, addressing learning bounds that were previously non-estimable from samples.

arXiv cs.AI / cs.LG / cs.CL·Sep 10, 17:57 UTC · 5d agoAI research1

28

Former TikTok execs built an app that uses AI to teach you how to pose for a photo

Ex-TikTok employees launched Superpose, an iOS camera app using generative AI to suggest photo poses, with $2.2M raised.

Former TikTok employees Melody Chu and Jing Liu launched Superpose, an iOS camera app that generates four AI pose suggestions per photo, with over 22,000 downloads and 190,000 poses generated since July. The app offers five free generations daily, with paid packs of five for $2.99 or 20 for $9.99. The startup raised $2.2 million from Khosla Ventures, Meitu, and OVTR VC, competing with Google's Camera Coach and Adobe's AI photo critique features.

TechCrunch · AI·Sep 15, 14:42 UTC · 1d agoAI industry

12

You Can't Prefer Emotions You Don't Sample: Intensity Undershoot in DPO-Tuned LLMs

Study quantifies DPO-tuned LLMs undershooting requested emotional intensity, tracing the gap to candidate-pool extremity rather than conditioning format.

Conditioning an instruction-tuned LLM on continuous valence-arousal targets yields gain of only 0.26 for valence and 0.13 for arousal on Llama-3.1-8B, far below faithful control of 1.0. The authors attribute undershoot to neutral-heavy preference corpora like EmoBank and candidate pools lacking extreme affect, leaving DPO without extreme exemplars. Uniform target coverage with a hotter candidate pool raises valence gain to 0.40 on Llama-3.1-8B and 0.44 on Qwen3-8B, with modest in-distribution cost; arousal gains remain unstable across seeds.

arXiv cs.AI / cs.LG / cs.CL·Sep 7, 17:47 UTC · 8d agoAI research

31

APT36 Suspected in PATCHCORD Espionage Campaign Using Google Sheets C2

Acronis links the PATCHCORD espionage campaign targeting Afghan telecom and South Asian critical infrastructure to APT36 with moderate confidence.

Acronis Threat Research Unit documented PATCHCORD, a previously undocumented C/C++ backdoor delivered via fake VPN installers impersonating Afghan Telecom (AFTEL), which hijacks Edge, Chrome and Firefox shortcuts for persistence and executes in-memory shellcode. A related Go implant, SHEETCORD, abuses the Google Sheets API for per-victim C2 and was distributed via a domain impersonating India's National Informatics Centre, while a third tool, HACKERAI C2 Agent, uses GitHub Gists and shows signs of LLM-assisted coding. An exposed staging server revealed SuperShell, RAT frameworks, credential-harvesting tools and OpenSSH exploit code, and the researchers attribute the activity to APT36/Transparent Tribe at moderate confidence.

Security Affairs·Aug 16, 07:24 UTC · Aug 16, 2026Threat actor in the wild

72

SynthGait-19K: A Physically Grounded Synthetic Video Dataset for Gait Parameter Estimation

Researchers release SynthGait-19K, a synthetic video dataset with 19,272 walking videos for training gait parameter estimation models.

SynthGait-19K is a physically grounded synthetic video dataset built from 6,427 MoCap sequences of 437 subjects, yielding 19,272 walking videos with SMPL motion and annotations for six gait parameters. The authors introduce Gait2Vid, a pipeline that unifies heterogeneous MoCap recordings and synthesizes RGB videos under controllable viewpoints, validating gait events against force-platform measurements. Using the dataset they benchmark direct RGB, pose-based, biomechanical, and human-mesh-recovery approaches, and introduce GaitXFormer as a direct RGB reference model. Findings show synthetic supervision transfers to real video, while spatial gait parameters are more sensitive to visual domain shift.

Hugging Face daily papers·Sep 7, 20:00 UTC · 8d agoAI research

28

Mistral AI raises 3 billion euros in Europe's largest-ever tech funding round despite lagging behind rivals

Mistral AI closed a €3 billion Series D led by Samsung Electronics, valuing the lab at over €21 billion.

Mistral AI raised €3 billion in what it calls the largest equity round ever by a European tech company, lifting its valuation past €21 billion after ASML valued it near €12 billion in September 2025. Samsung Electronics leads the round with EQT's Scaleup Europe Fund and PSG Equity as co-leads; new investors include Advent, BlackRock-managed funds, and Luxembourg. The company also took an $830 million loan in March to fund its own data centers and serves 125+ enterprise customers including Airbus, ASML, and HSBC. Mistral Medium 3.5 trails open rivals like Qwen and Kimi and doesn't compete with top closed US models.

The Decoder·Sep 8, 07:45 UTC · 8d agoAI industry1

65

One Symptom, Three Levers: A Critical Review of On-Policy Self-Distillation

A review paper frames on-policy self-distillation collapse as governed by three levers: token weighting, privileged information, and guidance decay.

The paper critically reviews On-Policy Self-Distillation (OPSD), where a language model trains on its own generations scored token-by-token by a teacher conditioned on privileged information such as reference solutions or environment feedback. It identifies collapse, the progressive narrowing of producible reasoning paths, as the dominant failure mode and analyzes it through three levers: signal weighting, the nature of privileged information, and teacher dynamics. The review is restricted to mathematical reasoning, reports no new experiments, and offers a shared vocabulary separating settled findings from disputed ones.

Hugging Face daily papers·Aug 25, 20:00 UTC · 21d agoAI research

45