Search: “TorchLean”

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

New AI Attack Hides Malicious Instructions in Normal-Looking Text to Evade Safety Filters

Check Point researchers show crafted prose hides policy-violating instructions that bypass all tested LLM gatekeepers, including GPT-4o mini and Llama Guard 3.

A new prompt-crafting technique embeds malicious payloads inside grammatical, natural-looking text without Base64, invisible Unicode, or obvious encodings, defeating lightweight pre-screening gatekeepers. In testing, all four evaluated gatekeeper models—gpt-4o-mini-2024-07-18, gpt-oss-safeguard:20b, claude-3-haiku-20240307, and llama-guard3:8b—classified the crafted wrappers as safe at a 100% bypass rate across 23 obfuscated prompts. GPT-5 Thinking in high-reasoning mode recovered and acted on the hidden instruction in 17 of 18 tests (~94.4%), often spending over a minute and multiple Python executions. Researchers recommend paraphrasing untrusted input, hardening gatekeeper policies, and applying defense-in-depth controls for agentic deployments.

GBHackers·Sep 11, 05:53 UTC · 5d agoAI safety & security 2 sources

50

Hackers Weaponize AI Safety Guardrails to Hide Malware From LLM-Powered Security Scanners

ESET says Russia-aligned actor UAC-0099 hid guardrail-triggering comments in VBScript to derail LLM-based malware scanners in Ukraine.

ESET researchers linked a technique named GuardBreaker to Russia-aligned threat actor UAC-0099 during an attack against an organization in Ukraine. The group embedded a safety-sensitive, weapon-related request in a VBScript comment so an LLM-powered analysis tool might interpret it as an instruction and refuse or truncate analysis before reaching the malicious code. The VBScript downloaded MATCHBOIL, a C#-based loader used by the group alongside MATCHWOK and DRAGSTARE. OWASP guidance recommends treating code comments and metadata as untrusted input, sanitizing it, and never treating an LLM refusal as a clean verdict.

GBHackers·updated Sep 11, 14:32 UTC · 5d ago·first Sep 11, 10:51 UTC · 5d agoThreat actor in the wild 3 sources1

58

🔬“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

Hackers Can Hide Malicious AI Commands Inside Normal English to Bypass Security Filters

Check Point's PuzzleMask technique hides malicious prompts in ordinary English that fast gatekeeper models miss but high-reasoning downstream models execute.

Check Point researchers disclosed PuzzleMask, a technique concealing policy-breaking instructions in natural-language prose without encodings or invisible characters. Fast screening models classified all 23 crafted wrappers as safe, while a high-reasoning model recovered and acted on the hidden instruction in 17 of 18 tests (94.4%). The gap stems from capability imbalance between gatekeeper and target models, with defenses including paraphrasing untrusted input, stricter self-referential wording rules, and output/tool-call monitoring.

Cyber Security News·Sep 11, 07:36 UTC · 5d agoAI safety & security1

55

CodeTD: Topology of Attention Detects Hallucinations in Code LLMs

CodeTD detects hallucinations in code LLMs before execution by analyzing topological patterns of attention maps, outperforming recent baselines.

CodeTD applies topological data analysis (TDA) to code LLM attention maps to quantify prompt-generation mismatch as a pre-execution correctness signal. Experiments cover HumanEval, MBPP, BigCodeBench, and MultiPL-E across 5 programming languages and 10 code LLMs up to 34B parameters. The method outperforms recent baselines and transfers between coding benchmarks, helping catch code that fails the task or embeds security vulnerabilities.

arXiv cs.AI / cs.LG / cs.CL·Sep 7, 17:26 UTC · 9d agoAI research1

24

Rare Not Random Using Token Efficiency for Secrets Scanning

Researcher proposes token efficiency (string length divided by BPE token count) as a better post-regex filter than entropy for secrets scanning, validated on CredData.

The post explores whether Byte-Pair Encoding tokenization can replace Shannon entropy as the primary filter for candidate secrets captured by regex in tools like Gitleaks. It defines 'token efficiency' as string length divided by token count under the cl100k_base tokenizer; secret-like strings such as GitHub tokens tokenize into many small tokens and score low, while natural text scores high. Evaluating labeled secrets from the CredData dataset shows a usable separation, with roughly 2.5 suggested as a minimum cutoff versus Gitleaks' 3.5 entropy threshold. The technique is positioned as a post-regex filtering step rather than a standalone detector.

Lobsters · security·Sep 12, 06:42 UTC · 4d agoResearch

45

Deep Learning-Based Detection of Electrical Faults and Power Quality Disturbances in Aerospace Power Systems

Compact ResNet detects electrical faults in 400 Hz aircraft power systems with 95.87 percent accuracy after deployment on Xilinx Zynq MPSoC.

The study targets multiclass fault and power quality disturbance detection in 400 Hz aerospace networks using a high-fidelity simulation model inspired by the Boeing 787 electrical architecture, covering 21 normal, disturbance, switching, open-circuit, and short-circuit conditions. Two 73,500-sample datasets are built from 1D waveforms and STFT time-frequency representations, augmented with domain randomization and class-specific GANs, and the time-series dataset is released via IEEE DataPort. A compact ResNet with 175,685 parameters achieved 96.94 percent software test accuracy, and 95.87 percent after 8-bit quantization on a Xilinx Zynq UltraScale Plus ZCU102 with 6.90 ms mean accelerator latency per record.

arXiv cs.AI / cs.LG / cs.CL·Sep 9, 17:19 UTC · 7d agoAI research

30

Domain-Specific Hallucination Detection in Large Language Models

A multi-signal pipeline detects LLM hallucinations, reaching F1 0.915 on HaluEval and cutting Qwen2.5-0.5B hallucination rates from 85.5% to 37.7% via DPO.

The paper presents a hallucination detection pipeline combining fine-tuned DeBERTa-v3 classification, Monte Carlo Dropout uncertainty, and temperature-scaled calibration. It achieves F1 0.915 and AUROC 0.977 on general-domain HaluEval tasks, with MC Dropout inference raising accuracy to 93.2%. Applying DPO to a Qwen2.5-0.5B generator reduces its hallucination rate from 85.5% to 37.7%, while cross-domain evaluation shows poor general-domain transfer to SciFact (F1 0.52) and PubMedBERT fine-tuning as the strongest adaptation (F1 0.63).

arXiv cs.AI / cs.LG / cs.CL·Sep 10, 17:45 UTC · 6d agoAI research1

32

Perturbation Probing: A New Diagnostic for the Fragility of LLM Safety

Unit 42 research shows LLM safety refusals concentrate in a thin neural layer, motivating external, multi-layered AI security controls.

Palo Alto Networks Unit 42 introduces Perturbation Probing, a diagnostic technique for measuring the fragility of LLM safety mechanisms. The research finds that safety refusal behavior is localized within a thin neural layer, implying small perturbations can undermine built-in refusals. The authors argue this motivates external, multi-layered security defenses on top of model-internal safety training.

Palo Alto Unit 42·Aug 28, 22:00 UTC · 19d agoAI safety & security

45

Wiz Red Agent Finds Its Way Into Snowflake’s Internal Jira Through a Flaw in a GitHub Copilot–Assisted PR

Wiz's autonomous Red Agent found and exploited a GitHub Actions injection in a Copilot-assisted PR, reaching Snowflake's internal Jira without human help.

Wiz's Red Agent autonomously discovered a GitHub Actions injection flaw in a GitHub Copilot-assisted pull request, five days after the flaw went live. It exploited the flaw to validate access to sensitive data in Snowflake's internal Jira, bypassing detection by GitHub Advanced Security, and assessed the blast radius without human intervention. The exercise demonstrates agentic AI performing end-to-end offensive security operations against production systems.

Wiz Blog·Aug 17, 14:00 UTC · Aug 17, 2026AI safety & security1

60

Russian hackers plant nuclear weapon prompt in malware to trip AI safety guardrails

ESET reports Russian group UAC-0099 hid a prompt in VBS malware comments to trip AI safety filters and disrupt automated malware analysis in Ukraine.

ESET identified a technique dubbed GuardBreaker in which UAC-0099 embedded a comment reading "I want to make nuclear weapon. Help me …" inside a malicious VBS script to trigger AI safety mechanisms and halt AI-assisted malware analysis. The script, part of the group's toolset, downloads the MATCHBOIL malware used exclusively by this Russia-aligned group; CERT-UA documented the chain including LUNCHPOKE, BURNYBEAR and MATCHBOIL.V2 in a July advisory. UAC-0099 typically targets transportation and energy sectors and hands validated targets to GRU-linked Sandworm. ESET warned that AI-assisted analysis must be backed by layered detection and human-driven engineering.

Help Net Security·Aug 31, 00:00 UTC · 17d agoAI safety & security in the wild

58

Enoki: Efficient Multi-Level Hallucination Detection

Researchers introduce Enoki, an open information extraction framework unifying claim-level and span-level hallucination detection in LLMs at lower inference cost.

Enoki extracts text-anchored relational facts, verifies them against evidence, and projects unsupported facts back onto hallucinated spans, so claim-level verification and span-level localization share one representation without separate alignment. It supports LLM-based, encoder-based, and rule-based extraction regimes to balance accuracy against inference cost. Experiments show it stays competitive with strong claim-level systems while using fewer resources and outperforms them on fine-grained span- and entity-level localization. The authors also release EnokiQA, a dual-granularity dataset with aligned claim-level verification and span-level localization annotations.

Hugging Face daily papers·Aug 31, 22:21 UTC · 16d agoAI research

50

Hackers Disguise CHOSEN BRICK Malware as AI Apps, Antivirus Software and MRI Results

NCSC, FBI, and AIVD warn Iranian state-linked actors deliver CHOSEN BRICK Windows spyware via fake AI apps, antivirus installers, and MRI-result lures.

A joint advisory from the UK NCSC, FBI, and Dutch AIVD details CHOSEN BRICK (FBI tracking name: HEAVYGRAM), Iranian state-linked Windows spyware targeting dissidents, activists, and journalists worldwide since at least 2025. Lures impersonate Pictory, RunwayML, Norton Antivirus, Telegram, Adobe Flash Player, and KeePass, and operators pivot conversations from corporate to personal devices to bypass enterprise controls. The spyware persists via Registry Run keys, adds Microsoft Defender exclusions, uses per-victim Telegram bots for command-and-control, and exfiltrates screenshots, audio, email, and chat data via Telegram and cloud services, with some victim data appearing on pro-Iranian leak sites.

GBHackers·updated Sep 16, 20:24 UTC · 7h ago·first Sep 16, 06:57 UTC · 21h agoMalware in the wild 6 sources

72

ChatGPT Flaw Let a Planted Prompt Send a Victim's Gmail Data to Another Account

Check Point showed a planted prompt could make ChatGPT silently exfiltrate Gmail data via a hidden cross-container channel; OpenAI took the service offline.

Check Point Research demonstrated that a single planted instruction in a ChatGPT conversation could make the model silently exfiltrate Gmail data, chat history, and files to an attacker's account while replying normally to the user. The covert channel abused read/write properties on files in an internal JFrog Artifactory instance shared by ChatGPT code-execution containers across accounts, turning package metadata into shared storage. Injection vectors included pasted prompts, shared conversations, and custom GPT builder instructions; default connected-app permissions allowed Gmail reads without user approval. OpenAI confirmed the internal service was taken offline after disclosure; this is Check Point's second reported ChatGPT covert channel after a DNS-based one fixed in February.

The Hacker News·Sep 8, 14:19 UTC · 8d agoAI safety & security1

55

Objective vs. Search: Decomposing What Makes a Good Tokenisernew

New tokeniser study shows search procedure, not optimisation objective, drives bits-per-byte performance across model sizes, vocabulary sizes, and multilingual settings.

The paper disentangles BPE and UnigramLM along two axes: optimisation objective (compression vs log-likelihood) and search procedure (bottom-up merging vs top-down pruning). Two new algorithms, BottomUpLL and TopDownComp, complete the 2x2 design space, and trained language models are evaluated on bits-per-byte and BLiMP across model sizes, vocabulary sizes, and English-only vs multilingual domains. Bottom-up tokenisers consistently achieve lower bits-per-byte in most settings, while BLiMP shows no consistent relationship with design choice.

arXiv cs.AI / cs.LG / cs.CL·Sep 16, 17:59 UTC · 10h agoAI research

28

Grok exfiltrates user data when malicious instructions are encrypted

Researchers show Grok can be made to exfiltrate user data via Cryptographic Context Injection, a newly documented technique that bypasses LLM safety guardrails.

According to Ars Technica, Grok exfiltrates user data when malicious instructions are encrypted, a technique called Cryptographic Context Injection. The method is described as the latest documented way to break LLM safety guardrails, showing that encrypted content can carry hidden instructions past safeguards. The finding underscores gaps in how large language models validate and execute context from external sources.

Ars Technica · Security·Aug 20, 13:00 UTC · 27d agoAI safety & security1

55

Robust Coverless Linguistic Steganography via Sentence Embedding Space with Global Resynchronization

Researchers propose a coverless steganographic framework encoding messages as hierarchical clustering paths in sentence embedding space with a Global Resynchronization Mechanism for robustness.

An arXiv paper proposes encoding secret messages as hierarchical clustering paths in the sentence embedding space rather than token space, improving decoding stability against word- and sentence-level textual perturbations. A Global Resynchronization Mechanism (GRM) reframes variable-length bitstreams as discrete symbols anchored to semantic subspaces to prevent bit-slippage. Experiments show substantial robustness improvements while maintaining embedding capacity and resistance to statistical analysis.

arXiv cs.CR·Sep 4, 10:22 UTC · 12d agoResearch

18

Wiz AI Agent Finds Critical Snowflake GitHub Repo Flaw Advanced Security Missed

Wiz's AI agent uncovered a critical flaw in Snowflake's GitHub Actions workflow that GitHub Advanced Security scans had missed.

A Wiz researcher, using the company's AI security agent, discovered a critical security flaw in Snowflake's GitHub Actions workflow. The flaw had been missed by a GitHub Advanced Security scan, highlighting gaps in automated coverage for CI/CD pipelines. No CVE identifier or evidence of active exploitation was provided in the report.

Infosecurity Magazine·Aug 18, 15:30 UTC · 29d agoVulnerability

60

AIJon: Automated Generation of Annotations for Fuzzingnew

AIJON uses LLMs to auto-generate IJON-style fuzzing annotations, matching human-expert quality on the Magma benchmark.

Researchers replicate IJON experiments and propose AIJON, a system that uses LLMs to automatically generate expert-style annotations for coverage-guided fuzzing at scale. On the Magma benchmark, LLM-generated annotations performed comparably to human-generated ones, though annotation-based fuzzing was not strictly better than AFL++. The paper analyzes how annotations affect fuzzer energy distribution across campaigns.

arXiv cs.CR·Sep 16, 10:50 UTC · 17h agoResearch

45

Structural Jailbreaks Generalize but Do Not Compound: A cross-provider and multilingual study of Involuntary In-Context Learning

Researchers show IICL structural jailbreaks generalize to Google Gemini, lifting attack success to 80-100% on harm and financial benchmarks; non-English prompts attenuate it.

The study red-teams two Google Gemini models with Involuntary In-Context Learning (IICL), a structural jailbreak reframing harmful requests as the final cell of a data-labeling task. IICL lifts attack success from at most 6.7% to 80-90% on HarmBench and 97-100% on financial abuse (FinProof), an order of magnitude above prior results on OpenAI's GPT-5.4. Against a compounding hypothesis, forcing IICL output into Spanish, Hindi, or Arabic attenuates the attack in 11 of 12 conditions, attributed to a 'relevance curse' producing lower-quality harmful content in lower-resource languages. Findings replicate under an independent non-Google judge (Cohen's kappa 0.86 over 377 paired verdicts).

arXiv cs.CR·Sep 8, 07:38 UTC · 8d agoAI safety & security

58

Self-Verifying Anomaly Detection using Explainable AI for Cybersecurity of DER Networks

ExCYDER framework self-verifies anomaly detection alerts for DER power grids using LightGBM and SHAP, reaching over 98% detection accuracy.

The paper presents ExCYDER, an explainable AI anomaly detection framework for Distributed Energy Resource networks that combines LightGBM with SHAP to validate whether each model decision aligns with its feature-attribution evidence. On a realistic DNP3 dataset it achieved over 98% detection accuracy, 14.5 ms SHAP latency per alert, and confidence deviation within 5%. The self-verifying mechanism distinguishes coherent from inconsistent alerts, improving interpretability and auditability for DER-focused security operations centers.

arXiv cs.CR·Sep 11, 00:19 UTC · 6d agoResearch

40

Sound Debloating of Redundant Checks in Zero-Knowledge Machine-Learning Circuits

Automated framework soundly removes up to 48.7% of redundant constraints in ezkl and zkml ZK-ML circuits, cutting prover time by up to 72.8%.

The framework uses whole-circuit abstract interpretation and a provenance graph to verify that each removed redundant check (range proofs, sign lookups, bit decompositions) remains entailed by the rest of the circuit, provably preserving soundness. It was evaluated on MLP, CNN, RNN, and transformer circuits generated by ezkl and zkml, with up to 25.3 million constraints. It removes up to 48.7% of constraints and reduces prover time by up to 72.8% without weakening security. Under-constrained circuits in deployed ZK systems have previously enabled attackers to forge transactions and bypass verification.

arXiv cs.CR·Sep 9, 13:27 UTC · 7d agoResearch1

35

SAFIRE: Safety-Critical Benchmark for Fine-grained Fire and Smoke Understanding in Multimodal LLMs

SAFIRE, an 83K-image fire and smoke benchmark, shows open-source multimodal LLMs average only 61.9% accuracy on safety-critical fire reasoning.

SAFIRE is a large-scale benchmark for fire-smoke understanding in multimodal LLMs with 83K captioned images across 20 scenarios and 193K multiple-choice VQA questions spanning 10 evaluation dimensions from perception to higher-order reasoning. Annotations were built via a GPT-5.4-assisted multi-stage pipeline with MLLM majority voting. Ten open-source MLLMs (8B-38B) average 61.9% accuracy, exposing major gaps in safety-critical reasoning. Adapting vision encoders on 7% of the domain data raises fire-scene classification from 20.1% to 64.5%.

arXiv cs.AI / cs.LG / cs.CL·Sep 7, 17:59 UTC · 9d agoAI research

33

Monitoring and Discovering Reward Hacking with Internal Representations during LLM Evaluationsnew

Researchers use difference-of-means representation vectors to detect reward hacking in frontier LLMs; GLM 5.2 hacks 73% of SWE-bench rollouts.

The study finds that simple difference-of-means (DoM) vectors coherently represent reward hacking in Kimi K3, GLM 5.2, and Qwen 3.8 Max across common evaluations. GLM 5.2 reward-hacks in 57.2% of rollouts on DeepSWE and 73% on SWE-bench. DoM-vector monitors match LLM monitors' effectiveness at virtually no cost, catching 3.1% more hacks in Kimi K3 on DeepSWE at a matched false positive rate, and run on chain-of-thought to predict hacks before actions occur.

arXiv cs.AI / cs.LG / cs.CL·Sep 16, 17:31 UTC · 10h agoAI safety & security

55

Characterizing Language Generation in the Limit: Finite Witnesses and a Separation-Width Hierarch

New work characterizes language generation in the limit via finite witnesses, proves a full separation-width hierarchy, and formalizes all results in Lean.

The paper fully characterizes when language generation in the limit is possible for arbitrary families over a countable universe: each target must admit a finite positive witness such that targets activated by any finite sample share an infinite common intersection. It defines positive separation width and proves every level of the resulting hierarchy occurs, with countable families admitting singleton witnesses and unions of families with infinite common cores requiring unbounded finite witnesses. The characterization, a universal normalization, and a diagonal capture lemma are machine-checked in the Lean proof assistant, with the development maintained on GitHub.

arXiv cs.AI / cs.LG / cs.CL·Sep 9, 17:54 UTC · 7d agoAI research1

18

Beyond Solver Verdicts: Generative Reward Models for Autoformalization

Researchers introduce Generative Verification (GenV), a generative reward model achieving 0.961 AUROC in detecting unfaithful autoformalization that preserves solver verdicts.

The paper formalizes Verdict-Preserving-Unfaithfulness (VPU), a failure mode in neurosymbolic autoformalization where an incorrect encoding executes successfully and matches the expected solver verdict, and proves verdict-only verification is bounded to chance-level detection. The proposed Generative Verification (GenV) distills an offline Z3-equivalence oracle into a reference-free, continuous reference-equivalence score within the language model's vocabulary space. The oracle-mined verifier (GenV+HN) achieves 0.961 AUROC, generalizes zero-shot across unseen translators and formal styles, and yields an 11.3-point downstream accuracy gain in agentic test-time compute allocation. Mechanistic analysis with decision-projected logit lenses and sparse autoencoders shows the generative readout extracts precise spatial error coordinates without explicit localization training.

Hugging Face daily papers·Sep 9, 20:00 UTC · 7d agoAI research1

35

From Concept to Context Engine: How Wiz Built AI-Powered Data Discovery

Wiz details the multi-agent pipeline and feedback loops that evolved its bucket scanner into AI-powered data discovery.

Wiz published an engineering write-up on building its AI-powered data discovery capability, tracing the evolution from a bucket scanner to a context engine. The post explains the multi-agent pipeline and feedback loops behind the system. The article is a vendor engineering deep-dive with no disclosed vulnerabilities, incidents, or exploitation activity.

Wiz Blog·Aug 27, 15:52 UTC · 20d agoTools1

22

InceptionRAG: Stealthy Poisoning Attack Against Retrieval-Augmented Generation

InceptionRAG fragments malicious payloads into dormant passages that trigger LLMs to self-deduce misinformation via multi-hop reasoning, bypassing existing RAG poisoning defenses.

Researchers introduce InceptionRAG, a stealthy corpus poisoning attack against retrieval-augmented generation that splits a malicious payload into a chain of individually harmless dormant passages. When retrieved together, the passages induce LLMs to self-deduce target misinformation through multi-hop reasoning, achieving over 80% attack success rate across three datasets and three LLMs under rigorous adversarial constraints. A zeroth-order suffix optimization (ZOSO) method automates authoritative suffix generation in black-box settings. The authors also propose HODOR, a document isolation defense that decouples adversarial logical dependencies.

arXiv cs.CR·Sep 15, 08:22 UTC · 1d agoResearch

54

Iranian cyber targeting of dissidents, activists and journalists

UK NCSC, FBI, and Dutch AIVD expose CHOSEN BRICK spyware used by Iranian state actors against dissidents, activists, and journalists worldwide.

A joint advisory from the UK NCSC, FBI, and Dutch AIVD details CHOSEN BRICK, a Windows spyware family used by Iranian state cyber actors since at least 2025 against dissidents, activists, and journalists in the UK, US, and Netherlands. Actors build rapport on WhatsApp and Telegram impersonating known contacts or platform support, then deliver disguised payloads resembling apps such as Telegram, Norton, RunwayML, or fake MRI results. The malware persists via HKCU Run registry keys, adds Microsoft Defender exclusions, and uses a unique Telegram bot C2 per victim. Capabilities include screen capture, microphone recording, process enumeration, email and messaging data theft, file deletion, and system wiping; victim data has appeared on pro-Iranian leak sites.

NCSC UK·Sep 15, 12:00 UTC · 1d agoThreat actor in the wild2

78