Search: “Compsec Lab”

ScienceBuddy: Recursive-in-Recursive Self-Improvement for Interactive Scientific Agents

ScienceBuddy released: interactive scientific agent workspace coupling harness evolution with model reinforcement learning for continual self-improvement across four scientific task families.

ScienceBuddy is an interactive scientific research workspace that turns researcher requests, feedback, and execution evidence into tasks and evaluation rubrics for continual learning. Its recursive-in-recursive self-improvement paradigm couples harness evolution with the model fixed (inner recursion) and model reinforcement learning under the improved harness (outer recursion). Case studies span four scientific task families covering researcher interaction, harness refinement, and model learning. The system is released as a research product at science-buddy.io.

Hugging Face daily papers·updated Sep 15, 17:55 UTC · 11h ago·first Sep 14, 20:00 UTC · 1d agoAI research 2 sources

Google Chrome 153 Released With Fixes for 42 Security Vulnerabilitiesnew

Google shipped Chrome 153 to the Stable channel fixing 42 vulnerabilities, including three critical flaws in WebGL, Internals, and Workers; no exploitation reported.

Google released Chrome 153 (153.0.8010.47/48) for Windows, macOS, and Linux, patching 42 security vulnerabilities including three rated critical: CVE-2026-91726 (out-of-bounds read in WebGL), CVE-2026-91721 (use-after-free in Internals), and CVE-2026-91749 (use-after-free in Workers). Twenty-eight fixes are rated high severity, covering use-after-free, type confusion, race condition, integer overflow, and authorization flaws across components like V8, Skia, DOM, ServiceWorker, PDF, and Extensions. Google's bulletin indicates no vulnerabilities are currently being exploited in the wild, and external researchers earned rewards up to $1,500 for reported issues. Enterprises are advised to verify fleet-wide deployment via browser-management consoles and enable automatic updates.

GBHackers·05:00 UTC · 40m agoVulnerabilityCVE-2026-91726CVE-2026-91721CVE-2026-91749+27 CVEs

Fake CVE-2023

A fake PoC for WinRAR RCE CVE-2023-40477 posted on GitHub actually deploys VenomRAT through a multi-step infection chain targeting researchers.

Four days after Zero Day Initiative publicly disclosed the WinRAR RCE vulnerability CVE-2023-40477 on August 17, 2023, an actor using the alias whalersplonk published a fake proof-of-concept on GitHub. The Python script actually repurposed public PoC code for GeoServer SQL injection CVE-2023-25157 and triggered an infection chain ending in VenomRAT. The README and an accompanying video lured users into running the script; the video drew over 100 plays. Unit 42 assesses the actor was opportunistic, targeting other miscreants adopting new vulnerabilities rather than researchers specifically.

Palo Alto Unit 42·Aug 17, 10:57 UTC · 29d agoMalwareCVE-2023-40477CVE-2023-25157

Google Chrome Zero-day Vulnerability Exploited in the Wild (CVE-2026-87491)

Google patched actively exploited Chrome V8 out-of-bounds write CVE-2026-87491; CISA added it to KEV with a September 23 patch deadline.

Google fixed 230 Chrome vulnerabilities in its stable channel update, including CVE-2026-87491, an out-of-bounds write in the V8 JavaScript engine being exploited in the wild. The flaw, reported by Jihyeon Jeong of Compsec Lab, Seoul National University, is the seventh Chrome zero-day patched by Google in 2026. It affects Chrome versions before 153.0.8010.36, and CISA added it to the Known Exploited Vulnerabilities Catalog urging patching before September 23, 2026. Users should upgrade to Chrome 153.0.8010.36 (Linux) or 153.0.8010.36/.37 (Windows/Mac); Qualys customers can detect vulnerable assets with QID 388666.

Qualys ThreatPROTECT·Sep 10, 05:14 UTC · 6d agoExploit / PoC in the wildCVE-2026-87491CVE-2026-2441CVE-2026-3909+4 CVEs

REVSTEALER ramps up: analysis of up-and-coming infostealer

Elastic Security Labs analyzes REVSTEALER, an emerging infostealer harvesting browser credentials, cryptocurrency wallets, and gaming accounts.

Elastic Security Labs published a deep-dive analysis of REVSTEALER, described as an up-and-coming infostealer. The malware targets browser data, cryptocurrency wallets, and gaming accounts, with credential harvesting as its core function. The report tracks the family's capabilities as it ramps up activity in the wild.

Elastic Security Labs·Sep 2, 00:00 UTC · 14d agoMalware in the wild

Google fixes yet another actively exploited Chrome zero-day (CVE-2026-87491)

Google patched actively exploited Chrome zero-day CVE-2026-87491, a V8 out-of-bounds write enabling sandboxed code execution, in Chrome 153.0.8010.36+.

Google fixed 230 Chrome vulnerabilities including CVE-2026-87491, a Medium-severity out-of-bounds write in the V8 JavaScript and WebAssembly engine that allows remote arbitrary code execution inside the sandbox via a crafted HTML page. Google confirmed an exploit exists in the wild; fixes shipped in Chrome 153.0.8010.36 and .37 for Windows and macOS and 153.0.8010.36 for Linux. The flaw was reported August 6, 2026 by Jihyeon Jeong of Seoul National University's Compsec Lab, earning a $2,500 bounty. It is the seventh actively exploited Chrome zero-day patched in 2026, following CVE-2026-2441, CVE-2026-3909, CVE-2026-3910, CVE-2026-5281, CVE-2026-11645 and CVE-2026-85046.

Help Net Security·Sep 9, 07:53 UTC · 6d agoExploit / PoC in the wildCVE-2026-87491CVE-2026-2441CVE-2026-3909+4 CVEs

Scores Alone Do Not Prove Discovery: The Discovery Certification Protocol for Auditing AI Research Agents

Researchers introduce the Discovery Certification Protocol, an auditable test framework that verifies whether AI research agents' claimed discoveries are genuine.

The Discovery Certification Protocol (DCP) converts AI research agents' discovery claims into executable recovery and feedback tests organized as gated audits. Controlled audits in SQLite optimization and virtual catalyst control produced zero recoveries in 96 episodes, with an upper bound of 0.0468. A deterministic, LLM-free verifier reproduces audit decisions from frozen evidence, giving AI research a common evidence language for outcomes, alternative routes, and feedback effects.

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

Chrome V8 Zero-Day Exploited in the Wild Enables Code Execution Inside Sandbox

Google patches actively exploited Chrome V8 zero-day CVE-2026-87491, an out-of-bounds write enabling sandboxed code execution via crafted HTML pages.

Google released Chrome 153.0.8010.36/.37 fixing 230 vulnerabilities, including CVE-2026-87491, an out-of-bounds write in the V8 engine that allowed remote code execution inside the sandbox via a crafted HTML page and is confirmed to be exploited in the wild. The flaw was reported on August 6, 2026 by Jihyeon Jeong of Compsec Lab, Seoul National University, who received a $2,500 bug bounty. This is the seventh actively exploited Chrome zero-day of 2026. The update also fixes five critical flaws in WebGL and Cast, plus a high WebPackaging use-after-free (CVE-2026-87639) credited to OpenAI Codex Security.

The Hacker News·Sep 9, 09:11 UTC · 6d agoExploit / PoC in the wildCVE-2026-87491CVE-2026-2441CVE-2026-3909+10 CVEs

Researchers Use Claude to Port Pre-Auth RCE Exploit From One PLC Model to Another

Forescout used Anthropic's Claude to port a working pre-auth RCE exploit for CVE-2021-31886 between WAGO PLCs, executing ARM shellcode on live hardware.

Forescout Vedere Labs used Claude interactively to port an RCE exploit for CVE-2021-31886, a CVSS 9.8 stack buffer overflow in the Nucleus FTP server's USER command, from a WAGO 750-852 to a WAGO 750-831 PLC, running attacker-supplied ARM shellcode. The final RCE stage cost $535.74 in API usage over 8 hours 32 minutes, and a follow-up attempt to build a C2 implant permanently bricked the device. CERT@VDE lists many Nucleus V1-based WAGO models as vulnerable with no updates available; Siemens plans no Nucleus NET remediation. The work follows a joint NSA/CISA/FBI/DOE/EPA advisory warning of AI-generated exploitation scripts targeting internet-exposed Siemens S7 PLCs.

The Hacker News·Sep 2, 07:47 UTC · 13d agoResearchCVE-2021-318862

Google Chrome 153 Update Fixes 42 Security Flaws, Including 3 Critical Ones

Google shipped Chrome 153 fixing 42 vulnerabilities, including three Critical use-after-free and out-of-bounds bugs, with no active exploitation reported.

Google's Chrome 153 Stable channel update (153.0.8010.47/.48 for Windows/macOS, 153.0.8010.47 for Linux) patches 42 vulnerabilities: three Critical, 27 High, ten Medium, and one Low. The Critical flaws are CVE-2026-91721 (use-after-free in Internals), CVE-2026-91749 (use-after-free in Workers), and CVE-2026-91726 (out-of-bounds read in WebGL). Google's bulletin states none of the patched issues are actively exploited, and detailed bug links remain restricted until most users receive the fixes. Bug bounty awards include $1,500 to Hafiizh for CVE-2026-91724 and $1,000 to Jihyeon Jeong of Seoul National University for CVE-2026-91728.

Cyber Security News·03:01 UTC · 2h agoVulnerabilityCVE-2026-91721CVE-2026-91749CVE-2026-91726+27 CVEs

Zscaler Agentic SOC combines AI agents with zero trust telemetry

Zscaler launched Agentic SOC, an AI-agent-driven security operations platform combining zero trust telemetry with frontier models from Anthropic and OpenAI.

Zscaler announced Agentic SOC, a security operations platform built around specialized AI agents for triage, root-cause investigation, verdict assignment, and automated threat containment. The platform pairs Zscaler's zero trust telemetry, drawn from roughly 750 billion daily transactions and a large decoy mesh network, with frontier models from Anthropic and OpenAI plus proprietary threat intelligence. It features closed-loop inline remediation that can isolate compromised users, block command-and-control traffic, and cut off lateral movement, alongside a context graph that correlates third-party data. Continuous threat hunting combines AI automation with human experts from Zscaler and Red Canary, and customer Maire Tecnimont is cited as an early adopter.

Help Net Security·Sep 9, 10:09 UTC · 6d agoTools

Chrome 153 Patches Seventh Zero-Day of 2026

Google patches Chrome V8 zero-day CVE-2026-87491, exploited in the wild, marking the seventh exploited Chrome zero-day of 2026, in an update fixing 230 flaws.

Google released Chrome 153 with fixes for 230 vulnerabilities, including CVE-2026-87491, a medium-severity out-of-bounds write in the V8 JavaScript and WebAssembly engine that Google says is being exploited in the wild. It is the seventh zero-day patched in Chrome in 2026, following CVE-2026-2441, CVE-2026-3909, CVE-2026-3910, CVE-2026-5281, CVE-2026-11645 and CVE-2026-85046. The update also fixes five critical bugs (four in WebGL, one in Cast), 41 high-severity defects, and more than 180 medium/low issues, rolling out as 153.0.8010.36/.37 for Windows/macOS and 153.0.8010.36 for Linux.

SecurityWeek·Sep 9, 09:45 UTC · 6d agoExploit / PoC in the wildCVE-2026-87491CVE-2026-2441CVE-2026-3909+4 CVEs

Dr. Claw: An AI Scientist Workspace for Vibe Research

Researchers release Dr. Claw, an open-source auditable workspace that wraps coding agents like Claude Code for end-to-end AI-assisted research workflows.

Paper 2609.00365 presents Dr. Claw, an open-source workspace that wraps existing coding-agent executors such as Claude Code and Gemini CLI in a controllable, human-in-the-loop research workflow. It uses persistent state objects, a reusable skill library, and multi-executor coordination to make research decisions auditable and recoverable, rather than adding another autonomous agent. Holding the executor fixed, Dr. Claw scores higher on research completeness than a bare command-line agent while preserving an auditable process trail. The code is released under AGPL-3.0 on GitHub (OpenLAIR/dr-claw).

Hugging Face daily papers·Aug 30, 20:00 UTC · 16d agoAI research

Chrome 153 Fixes 230 Vulnerabilities, Including One 0-Day Exploited in the Wild

Google shipped Chrome 153 with 230 fixes, including CVE-2026-87491, a V8 out-of-bounds write zero-day exploited in the wild.

Chrome 153 (153.0.8010.36/.37) rolls out to Windows, Mac, and Linux with 230 security fixes, among the largest patch batches in recent Chrome history. The headline flaw is CVE-2026-87491, a Medium-severity out-of-bounds write in the V8 JavaScript and WebAssembly engine that Google confirmed is exploited in the wild; it was reported by Jihyeon Jeong of Compsec Lab at Seoul National University for a $2,500 bounty. The release also closes five Critical-rated flaws, including CVE-2026-87464, CVE-2026-87488, CVE-2026-87438, CVE-2026-87527, and CVE-2026-87628, mostly use-after-free and out-of-bounds write bugs in WebGL and Cast, plus 43 High-severity issues across ANGLE, PDFium, V8, DevTools, and Payments. Several bugs were surfaced with AI-assisted discovery tools, including OpenAI's Codex Security team, and top bounties reached $5,000 for CVE-2026-87504.

Cyber Security News·Sep 9, 04:46 UTC · 7d agoVulnerability in the wildCVE-2026-87491CVE-2026-87464CVE-2026-87488+7 CVEs1

Researcher Releases FalconFlank PoC Showing Privilege Escalation in CrowdStrike Falcon

Researcher Chaotic Eclipse released FalconFlank, a zero-day privilege escalation PoC abusing CrowdStrike Falcon's Office macro remediation; CrowdStrike is investigating and issued a tech alert.

Researcher Chaotic Eclipse published FalconFlank, a proof-of-concept zero-day local privilege escalation in CrowdStrike Falcon Sensor that abuses the Microsoft Office malicious macros remediation feature, tested on fully updated Windows 11 25H2 and Windows Server 2025. CrowdStrike said it is investigating, advised disabling the Microsoft Office File Suspicious Macro Removal policy, and pointed customers to a FalconFlank tech alert, noting cloud anti-malware protection remains active. The article also covers related PoCs from the same researcher: HardBreacher against Kaspersky (fixed via update), ShieldBreak (CVE-2026-69414) against Microsoft Defender, an unpatched patch bypass of CVE-2026-50656, GreenSection against NVIDIA, and PrettyPrague against Avast, which Gen Digital patched in versions 26.7.11086 and 26.8.11125.

The Hacker News·Sep 3, 06:26 UTC · 12d agoExploit / PoCCVE-2026-69414CVE-2026-50656

Researcher Drops New Microsoft Defender PoC Showing ShieldBreak Patch Can Be Bypassed

Researcher Chaotic Eclipse released a PoC showing CVE-2026-69414's patch is bypassable, allowing arbitrary file reads as SYSTEM on current Windows.

The researcher known as Chaotic Eclipse published a proof-of-concept for a zero-day in Microsoft Defender, dubbed ShieldCrash, assessed as a patch bypass for ShieldBreak (CVE-2026-69414, CVSS 7.8). The PoC demonstrates an arbitrary file read as SYSTEM with the latest Windows installed, and all supported desktop versions are said to be impacted. Microsoft patched the original issue in Microsoft Malware Protection Engine 1.1.26080.3, which updates automatically. The same researcher recently released PoCs for flaws in CrowdStrike Falcon Sensor, Kaspersky, Avast Antivirus and NVIDIA.

The Hacker News·updated Sep 10, 12:18 UTC · 5d ago·first Sep 9, 06:47 UTC · 6d agoExploit / PoC 8 sourcesCVE-2026-694141

Testing race conditions with memory access tracing and stack-based delay injection

Google Project Zero released MAccConc, Linux kernel tooling that traces memory accesses to explore and test race condition interleavings.

A Google Project Zero researcher published MAccConc (Memory Access Concurrency), tooling for exploring possible interleavings of multithreaded test cases in the Linux kernel, available on GitHub. The tools use KCOV with ASAN outline-mode instrumentation to record per-access memory traces, enabling automatic testing of all A-B-A interleavings plus terminal and GUI explorers for manual analysis. The work targets confirming race condition candidates, building reliable regression tests, and enabling concurrency fuzzing, drawing on ideas from SKI and Ned Williamson's sockfuzzer.

Google Project Zero·Sep 8, 07:00 UTC · 7d agoResearch1

ExecCritic: Learn to Test, Test to Improve for Coding Agents

ExecCritic separates test generation from patching for coding agents, lifting SWE-bench Verified resolution to 72.6%.

ExecCritic pairs a test-verify-revise scaffold with role-specific reinforcement learning: a Test agent writes repository-native tests and a Repair agent fixes code from execution feedback, both using Qwen-3.5-35B-A3B backbones. Post-trained Qwen agents compose to 72.6% on SWE-bench Verified, an 11.4-point gain over the 61.2% no-test baseline, without stronger-model or oracle feedback at evaluation time. The work shows test quality is the key variable: base-agent tests lowered resolution to 57.3% while GPT-5.6-sol tests raised it to 65.3%.

arXiv cs.AI / cs.LG / cs.CL·Sep 8, 17:53 UTC · 7d agoAI research1

NovaFabric: Tamper-Evident, Replayable Evidence for Autonomous AI Agent Runs

NovaFabric seals autonomous AI agent runs into tamper-evident, replayable Run Capsules enabling third-party audit under EU AI Act and ISO 42001.

NovaFabric records autonomous agent runs without modifying agent logic into portable Run Capsules (fifteen-entity schema) sealed with DSSE signatures, RFC 3161 timestamps, a Merkle log, and redaction attestations, supporting four-mode replay and third-party Evidence Bundle verification. Evaluation shows tampering rejected across three tested classes, 14/14 credential types redacted while preserving 9/9 decoys, and 140/140 mutations localized; blast-radius queries reach 45.5ms p99 over 10M edges. Limits include only 2/10 tool-using workloads completing replay due to missing tool-response substitution and ingest capped at 61.6 req/s. The contribution integrates OpenTelemetry, DSSE/in-toto, and W3C PROV rather than new cryptography.

arXiv cs.CR·Sep 11, 08:35 UTC · 4d agoAI safety & security1

Schrödinger's backup: not actually recovered until you try to restore IT

The Register sponsored feature urges admins to verify backups through restore testing before an incident happens.

A sponsored feature from The Register argues that backups cannot be considered recovered until restore tests succeed. It promotes verification practices and vendor tooling for disaster-recovery readiness.

The Register · Security·Aug 27, 15:00 UTC · 19d agoIndustry

What Fal.Con 2026 Reinforced: AI Makes Proving Exposure More Important Than Ever

Horizon3's Fal.Con 2026 recap argues AI-accelerated vulnerability discovery makes continuous attacker-based exposure validation essential for defenders.

In a Fal.Con 2026 recap, Horizon3 argues that AI is compressing the time between vulnerability discovery and exploitation, making attacker-derived evidence about real exploitability the key prioritization signal. Horizon3 announced it joined CrowdStrike's Project QuiltWorks, with NodeZero exploitability intelligence flowing into Falcon Next-Gen SIEM and Falcon Fusion SOAR workflows able to trigger NodeZero 1-Click Verify for remediation testing. The company reported running over 1,200 NodeZero demos during the show, and CrowdStrike CEO George Kurtz's keynote framed AI red teaming and offense-informing-defense as central themes.

Horizon3.ai·Sep 4, 17:46 UTC · 11d agoIndustry

Chaotic Eclipse Released A PoC For NVIDIA GreenSection Memory Corruption Zero-Day

Researcher Chaotic Eclipse released GreenSection, a PoC exploit for an NVIDIA Windows user-mode shared-memory flaw enabling out-of-bounds writes and potential dwm.exe compromise.

Researcher Chaotic Eclipse (aka Nightmare Eclipse) released a PoC named GreenSection for a zero-day in NVIDIA's Windows user-mode components. Multiple NVIDIA components share a global memory section in BaseNamedObjects with full read/write access to all users; although checks prevent misuse, runtime reuse of section data can cause an out-of-bounds write. The unstable PoC crashes applications using Vulkan or OpenGL and could potentially cross user boundaries or compromise the Windows Desktop Window Manager (dwm.exe), though impact was not fully investigated. The researcher recently published similar zero-day PoCs against Kaspersky Endpoint Security, Avast Antivirus, and CrowdStrike Falcon.

Security Affairs·Sep 7, 13:42 UTC · 8d agoExploit / PoC 2 sources

The VMs Powering Mobile Agents (Instinct, Claude Code)

A teardown reveals Claude Code runs in Firecracker microVMs with a Rust PID 1 and MITM'd egress, while Instinct rents E2B sandboxes with git-based memory.

The author inspects the virtual machines hosting cloud agents: Claude Code runs in a Firecracker microVM with a custom Rust init (process_api) as PID 1, a 324 MB Bun harness on a read-only disk, and 443-only MITM'd SSE egress to api.anthropic.com with host-rotated OAuth tokens and no inbound access. Instinct rents E2B sandbox-as-a-service Firecracker microVMs (Ubuntu 22.04, 2 vCPU, 1.9 GB RAM) where agent memory is a git repo of Markdown committed by the agent and pushed to S3 as a single bundle, using short-lived STS credentials. Both platforms rely on Firecracker, differing mainly in fleet operator and guest boot configuration.

Hacker News · AI·Sep 8, 04:36 UTC · 8d agoAI tools & infraHN 58↑ · 19 comments1

$536 and 8 Hours: AI Learns to Attack a Different PLC

Forescout used Claude to port a WAGO PLC exploit for $536 in 8 hours, and an AI-generated payload later permanently bricked the test device.

Forescout researchers used Claude Code with Ghidra, terminal access, and physical hardware to port CVE-2021-31886, a pre-authentication buffer overflow in the Nucleus FTP server, from a WAGO 750-852 PLC to the WAGO 750-831 without source code or debugger access. The final RCE development stage cost $535.74 in API fees over an 8-hour, 32-minute session (2.6k input and 1.3M output tokens) and required heavy researcher guidance. Claude produced working ICMP and UDP beacon exploits in 12 minutes after identifying that normal FTP command completion was overwriting its shellcode, fixed by omitting the CRLF terminator. A follow-up session extending the payloads toward a C2 implant wrote to flash-mapped memory and permanently bricked the PLC, underscoring the risks of autonomous agents operating against physical infrastructure.

Security Affairs·Sep 2, 13:50 UTC · 13d agoResearchCVE-2021-318861

Tuskira expands exposure management with Agentic Control Plane

Help Net Security·Aug 5, 00:00 UTC · Aug 5, 2026Exploit / PoC

Cross-Model Agreement as a Deployment-Time Reliability Signal for Automatic Polyp Segmentation

Referee-Based Quality Estimation flags unreliable polyp segmentations at inference without ground truth, reaching ROC-AUC 0.960 with SegFormer-B0 referees.

RBQE measures agreement between a primary segmentation model and an independently trained referee on a 1,223-image external benchmark drawn from four public datasets. A cross-architecture SegFormer-B0 referee achieves the strongest signal (ROC-AUC 0.960), beating a Test-Time Augmentation baseline by 0.055 ROC-AUC under an identical protocol. Excluding trivially separable empty-mask cases, ROC-AUC falls to 0.876 (SegFormer-B0) and 0.783 (same-architecture control), but RBQE's margin over baselines widens. Progressive rejection of low-agreement predictions increases mean Dice of retained outputs, supporting selective prediction at the cost of one extra forward pass.

arXiv cs.AI / cs.LG / cs.CL·Sep 9, 17:32 UTC · 6d agoAI research

Fake Claude Opus 5 app delivers malware and wipes its own tracks

A fake 'Claude Opus 5' GitHub repository impersonating Anthropic distributes RevStealer, a Windows infostealer stealing passwords, crypto wallets and VPN credentials.

Morphisec identified a malicious GitHub repository, Claude-Opus-5-Free-Desktop, impersonating Anthropic to distribute a roughly 101 MB download that launches RevStealer, a Windows information-stealing malware. The malware targets browser credentials, Windows Credential Manager, a dozen password managers, more than fifty cryptocurrency wallets, VPN, messaging, game launcher and streaming accounts, streaming stolen data to its server before deleting itself. It evades analysis with ten sandbox and VM checks, indirect system calls, Windows Defender exclusions, a fake verification window, and a Polygon blockchain smart contract fallback for C2. Only 1 of 66 antivirus engines flagged the sample at check time.

Help Net Security·Sep 1, 00:00 UTC · 15d agoMalware

HypoEvolve: Genetic Algorithms Enable Multi-Agent LLMs to Discover Scientific Hypotheses

HypoEvolve couples a generational genetic algorithm with specialized LLM agents to generate drug-repurposing hypotheses, beating six baselines on DepMap selectivity (0.171 vs 0.115).

HypoEvolve coordinates specialized LLM agents through a generational genetic algorithm in which scientific judgments and new proposals reshape a hypothesis population. Evaluation centers on drug repurposing, linking mechanistic explanations to target-level biological claims assessed via external measures adapted from DepMap and Open Targets. Across 34 cancer types, HypoEvolve scores highest against six baselines on both measures, with DepMap selectivity of 0.171 versus 0.115 for the strongest baseline, and gains generalize to held-out cancer types.

arXiv cs.AI / cs.LG / cs.CL·Sep 14, 17:44 UTC · 1d agoAI research1

Do AI Coding Assistants Check Before They Install? A Pre-Registered Demand-Side Audit of Trust Signals in the Research Software Supply Chain

Pre-registered audit finds AI coding assistants verified provenance signals in only 9 of 1,920 trials before installing research software packages.

The study tested whether AI coding assistants check machine-readable trust signals such as SBOMs, signed releases, and provenance attestations before installing six open-source research software projects spanning HPC and quantum computing. Three models under two operating modes produced 1,920 registered trials scored from container logs. Provenance signals were opened in only 9 of 1,920 trials (0.5%) and zero of 384 control trials, with no trial running a verification command. The authors conclude publishing signals is insufficient and verification must be built into the program running the assistant.

arXiv cs.AI / cs.LG / cs.CL·Sep 7, 16:55 UTC · 8d agoAI safety & security

How Far Can Synthetic Data Take Thai OCR?

Synthetic-only training adapts PaddleOCR-VL into Wayu-Paxa-OCR-Zero, cutting Thai printed-page CER from 6.64% to 1.24% without real labels.

The study disentangles which factors of synthetic OCR data transfer to real Thai documents, finding typeface diversity, 2D structure, and real handwriting glyphs matter most. Using 45,723 synthetic pages, the authors adapt the 0.9B-parameter PaddleOCR-VL-1.6 into Wayu-Paxa-OCR-Zero, reducing median CER from 6.64% to 1.24% on printed pages and from 74.87% to 20.55% on handwriting. The model outperforms Typhoon OCR v1 7B on all five evaluation sets.

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