Alpine Linux rebuilt its GitLab server and reset all passwords after attackers tried to exploit CVE-2026-97229, a maximum-severity integer overflow in the Onigmo regex engine bundled with Ruby, on October 20-21. Eight worker crashes occurred and four were partially successful. Missing logs and outbound connection records meant compromise could not be ruled out. Alpine had run GitLab 19.3.4, then 19.3.6, which lacked the Ruby patch; the fixed image is alpinelinux/gitlab:19.3.6 with Ruby 3.3.13.
GitLabAlpine LinuxDocker+1
CVE-2026-97229integer overflowOnigmo
Supply Chain, Blue Team & Detection, DevSecOpsยท Published and added 1 Nov 2026
JADE (Joint Action DEnoiser) is a diffusion-based vision-language-action architecture in which the VLM reads the current noisy action latents together with images and instructions at every denoising step, and guides the action expert through gated cross-attention. The VLM is also trained to recover clean action latents, and an adaptive step-size schedule cuts sampling steps. On RoboDojo it averages 12.67% success, the highest among ฯ0.5-based methods, and it adds 20.8 points over ฯ0.5 on four real bimanual tasks.
vision-language-action modelvision-language modeldiffusion model
JADEฯ0.5RoboDojo+2
research, multimodalยท Added 3 Nov 2026, published 31 Oct 2026
The paper audits six situational-safety benchmarks for multimodal LLMs and proposes the Benchmark Card, seven measured fields that show whether a SAFE/UNSAFE label score comes from the scene, the request or the benchmark files. MSSBench chat leaks its label through image size, which alone reaches 82.3% accuracy. The Spot-Stop decomposition splits ranking-versus-labelling gaps across fourteen open MLLMs, and label-free recentering improves ranking on VSCBench for 13 of 14 models.
vision-language modelLLM
MSSBenchMOSSBenchHoliSafe+2
evals, ai-safety, multimodal, researchยท Added 3 Nov 2026, published 31 Oct 2026
Hyperflow H3 is Video Rebirth's 8-step LoRA for MiniMax H3, released in September 2026 with open weights. Artificial Analysis ranks it #5 on AA-Video-I2V v2.1 and #3 on AA-Video-Editing v1.1, ahead of MiniMax H3 (768p). It costs $9.00 per minute for Image to Video and $18.00 for Video Editing at 1088p, under half of Dreamina Seedance 2.5 ($34.12 and $40.82). The thread gives per-capability and per-use-case breakdowns.
LoRAdiffusion model
Artificial AnalysisVideo RebirthHyperflow H3+6
models, evals, multimodalยท Added 31 Oct 2026, published 30 Oct 2026
Oversecured Apps Care, Part 1 (21 March 2024) publishes scanner results for about 224 Google Android apps, drawn from a wider scan of over 6,000 public APKs from 1,000+ companies. Tables per device image (phone, Wear OS, Android Automotive, Android TV) give package names, three finding counts and report links, including android, com.android.settings, gms and Play Store. It invites researchers to report issues to Google for rewards.
Oversecured guide to SDK security in mobile apps, built around Microsoft's April 2026 disclosure of an intent redirection flaw in EngageLab's EngageSDK (50 million installs, fixed in 5.2.1). It covers SDK attack surfaces (exported components, WebViews, deep links, hardcoded secrets, payment crypto), a canhub image cropper file-overwrite example, why source-only SAST misses SDK code, and MASVS and PCI DSS expectations.
RADIOlogy is a training-free, single-pass zero-shot anomaly detection method that uses multi-head attention pooling in vision-language models. It scores anomalies by contrasting average similarity between vision features and sets of normal and abnormal prompts, with no windowed execution. RADIO agglomerative features drive a dense refinement pipeline for pixel-level segmentation, and the method is applied to medical images.
CLIPRADIOvision-language model
RADIOlogyarXiv
research, multimodalยท Added 31 Oct 2026, published 29 Oct 2026
PhysCount is a benchmark of 546 real-world images and 4,595 human-written questions in 651 target-specific groups. It tests whether multimodal LLMs can separate visible appearances (direct view, reflections, glass) from the physical entities behind them. Across 40 MLLMs, counting appearances is much easier than linking them to entities, and the best model answers every question correctly in only 13.36% of groups.
vision-language model
PhysCountarXiv
evals, multimodal, researchยท Added 31 Oct 2026, published 29 Oct 2026
inkGround is a benchmark for stroke trajectory recovery from full-page, camera-captured handwriting, with 537 page images (about 18.7k words, 163k strokes) and pseudo-label references filtered by OCR agreement and human raters. Models get a page image and a target word, then output ordered stroke points. Seven VLMs were scored with DTW, Chamfer distance, readability and style similarity. Closed-source models, notably Gemini-3.1-Pro, did best, and the ranking held on IAM-OnDB.
vision-language model
inkGroundIAM-OnDBGemini-3.1-Pro+1
evals, multimodal, researchยท Added 30 Oct 2026, published 28 Oct 2026
Beyond the Event Horizon of Keccak extends a 2011 observation that a path in the Keccak sponge from a zero-capacity state back to one yields infinite second-preimage families. Three theorems cover loops from any capacity value x, k-block loops and cycles, and Joux-style multicollisions with search cost independent of k. Verified examples run on Keccak-f[200] and Keccak-f[1600] with c=32; it also proposes bridge-based protocols.
CISA alert of 27 October 2026 adding two vulnerabilities to the Known Exploited Vulnerabilities (KEV) Catalog on evidence of active exploitation: CVE-2026-48356, an unrestricted upload of a dangerous file type in Adobe Commerce and Magento, and CVE-2026-111497, an unchecked return value in Google Brotli. It restates BOD 26-04, which makes federal civilian agencies prioritise KEV flaws on exposed assets and check for prior compromise. No technical detail or IOCs are given.
Adobe CommerceMagentoBrotli
CVE-2026-48356CVE-2026-111497CISA KEV+2
AppSec, GRCยท Added 29 Oct 2026, published 27 Oct 2026
Know, then Draw is a learning paradigm for visual generation that uses autoregressive pretraining to acquire semantic knowledge and diffusion post-training for high-fidelity rendering. Controlled comparisons find autoregression the stronger pretraining objective, and semantic quality matters more than AR-only generation quality. This motivates a purely semantic tokenizer trained with language supervision only. A 3B text-to-image model trained in 6,600 H100 GPU-hours scores 86 on GenEval and 87 on DPG-Bench.
diffusion modelTransformer
arXivGenEvalDPG-Bench+1
research, multimodal, trainingยท Added 28 Oct 2026, published 27 Oct 2026
This paper presents an annotated egocentric tennis dataset (egocentric video synced with five allocentric views, global positions, contact times, stroke types) and evaluates image encoders, video encoders and vision-language models on seven prediction tasks. Best models reach 0.14 m player and 0.54 m ball localization error and 0.14 s contact-time error. Depth input helps 3D ball localization, and longer histories do not help forecasting. A conditional diffusion model generates whole-body tennis motion in real time.
vision-language modeldiffusion modelTransformer
arXiv
research, multimodal, evalsยท Added 28 Oct 2026, published 27 Oct 2026
This paper studies how multimodal LLMs refuse questions that an image cannot answer, using the MoHoBench benchmark on three MLLMs. It varies the meaning and placement of refusal instructions, compares supervised and preference-based post-training, and intervenes on a refusal-related activation direction. Instructions change refusal at the sample level in ways aggregate rates hide. Post-training raises refusal most but can mix refusal with attempted answers. Activation intervention gives smaller, transferable changes.
vision-language modelLLM
MoHoBencharXiv
ai-safety, multimodal, researchยท Added 28 Oct 2026, published 27 Oct 2026
Hi3D v1.0 (formerly Sparc3D) is an open-sourced image-to-3D model from Hi3D, which runs a closed v3.0 refining at 2048 cubes per side. In a test with 13 professionals in 3D algorithms, 562 judgments against eight other image-to-3D models favoured it 83.5% of the time. The thread also covers the Hi3D Agent and an MCP server for other agents. Back and bottom views were checked, where TRELLIS.2 collapsed a winged figure.
Model Context Protocol
Hi3DHi3D v1.0Sparc3D+2
models, multimodal, mcpยท Added 28 Oct 2026, published 27 Oct 2026
Aikido Images are free patched PHP container images on public ECR (public.ecr.aws/aikido-dev/php) that backport the PHP2Shell fixes for CVE-2026-109618 and CVE-2026-109619 into release lines 5.5 to 8.5, except 8.0. Upstream fixed 8.2 to 8.5 in October 2026 and left 8.1 and older unpatched. The post shows docker pull and Dockerfile FROM changes, then rebuild, restart, session.upload_progress.enabled=0 and compromise checks.
PHPDockerAikido Security+1
CVE-2026-109618CVE-2026-109619PHP2Shell
DevSecOps, AppSecยท Published and added 26 Oct 2026
A SecurityWeek report on a JavaScript loader campaign, tracked by Sansec since July 2025, that has infected 953 online stores, mostly Magento. The actor rotated through five loader domains, the latest being deepveindrilling.com (October 2026) behind Cloudflare, with visitor fingerprinting to hide the payload. Entry points include PolyShell (CVE-2026-48356) and SessionReaper (CVE-2025-54236), and a fully patched store was also infected. Sansec lists cleanup steps.
MagentoAdobe CommerceCloudflare+1
CVE-2026-48356CVE-2025-54236Sansec+10
Threat Intel, AppSecยท Published and added 26 Oct 2026
RoSA is a finetuning framework from Qiuyu Kong and three co-authors for long-text image-text retrieval. It targets "saliency collapse", where models over-rely on text tied to prominent visual features and miss contextual information. RoSA splits pairs into salient and contextual views using object-level saliency estimation and LLM text decomposition, then trains on both with plug-and-play auxiliary alignment objectives. Code and checkpoints are public.
LLM
RoSAQiuyu Kong
multimodal, research, trainingยท Published and added 26 Oct 2026
A dermatologist tested GPT-6 Astra and Claude Fable 5.1 on 200 biopsy-proven lesions from MILK10k, giving each a clinical and a dermoscopic image. A CNN (Inception v3, fine-tuned for about twenty minutes on 68,011 ISIC images) named the right diagnosis for 62% of lesions, against 42% for both LLMs. The post lists caveats (small sample, single run, possible contamination) and links the cases, prompts, replies and scoring code.
Falko Strenzke's ePrint 2026/2635 presents a multi-target second-preimage attack on message hashing in HSS/LMS and XMSS with SHA-256 and SHA-256/192. It herds chaining states from many collected signatures into one state, defeating the key and leaf-index prefix. With 2^64 signatures, SHA-256 cost falls from 2^256 to about 2^166 (2^156 at maximal message length). SHAKE256 parameter sets are unaffected.
Hash-only rollup: an OP Stack L2 (chain ID 42069000, live since 30 September 2026) where accounts are keccak-256 hash chains and transactions carry no signatures. Spending uses a commit-reveal step through two new transaction types in op-geth. The post describes a flaw the authors found: unauthenticated commits could be overwritten, which allowed preimage front-running. It was fixed with a first-commit-wins rule on 21 October 2026. It lists the remaining limits: L1 ECDSA, no fault proofs, a centralized sequencer, 160-bit addresses and wallet desync.
Corgi is a terminal multiplexer for humans and AI coding agents, installed with a curl script. It shows each agent's state (working, needs you, finished, idle), keeps agents running in the background like tmux, and tracks token usage, per-workspace resource use and every file an agent changed. It also supports git worktrees, SSH image paste (needs Ghostty Nightly) and a corgi --skill command that teaches agents to drive it.
CorgiGhosttytmux+1
coding-agents, agentsยท Added 24 Oct 2026, published 23 Oct 2026
Microsoft's analysis of the LiteLLM 1.107.0 compromise, published to PyPI and container images on 21 October 2026. The wheel shipped cook.pth, cook.py and sitecustomize.py, so a credential collector (PRIVCOLLECT) ran whenever Python started, even without importing LiteLLM. It covers AES/RSA-encrypted exfiltration to litellm.dev, a re-armable Unix loader, Defender KQL hunting queries, ATT&CK mapping and rotation guidance.
LiteLLMPythonPyPI+3
cook.pthPRIVCOLLECTT1195.001+4
Supply Chain, Blue Team & Detection, Threat Intelยท Published and added 23 Oct 2026
OWLS (Open-world Language-conditioned Speed Fields) plans directly over a vision-language model's dense feature map. A text prompt describing traversable terrain scores every pixel of one camera image. The scores are projected onto the ground, adapted with two morphological operations, and used as the speed field of a minimum-time planner. It needs no training, terrain labels, map or extra sensors. It was tested on a physical rover, on simulated Mars and Moon rovers over 200 procedural terrains, and on real planetary imagery.
vision-language model
OWLSarXiv
research, multimodalยท Added 26 Oct 2026, published 23 Oct 2026
DeepFaune-UK is a fine-tune of the open DeepFaune camera-trap classifier on 28 UK classes, paired with MegaDetector v6. On 20,212 MammalWeb images it reached 88.8% strict accuracy against 56.5% (DeepFaune v1.5.0) and 59.0% (SpeciesNet). Re-applying the recipe to the DINOv3-based classifier gave 90.6%. Fine-tuning on about 150 crops per class matched full-data out-of-distribution accuracy, and a DINOv2 run took about 22 minutes on under 4 GB of GPU memory.
DeepFauneMegaDetectorSpeciesNet+2
DeepFaune-UKMammalWeb
research, training, multimodalยท Added 26 Oct 2026, published 23 Oct 2026
UAV-SwarmAgent is a hierarchical multi-agent system where an operator gives one natural-language sentence and a ground-station agent writes and runs code that produces per-UAV routes. A vision-language model grounds targets in an aerial map, and worker agents verify incidents with a compact onboard model. In simulation it located the incident in 244 of 255 missions (95.7%), and flight time followed an Amdahl-type law with speedup bounded at 1.52x for one to five UAVs.
vision-language modelLLM
UAV-SwarmAgentarXiv
research, agents, multimodalยท Added 26 Oct 2026, published 23 Oct 2026
PHP2Shell is Calif's writeup of pre-auth RCE in PHP (CVE-2026-109618, CVE-2026-109619) triggered by memory_limit fatal errors that leave dangling state. It covers a HashTable bucket bug in _zend_hash_str_add_or_update_i, a Session upload-progress stale-pointer bug, and an unserialize destructor-list primitive that reaches PHP-FPM. It lists affected apps (WordPress, Drupal, Joomla, Magento and others) and gives temporary mitigations.
Decode-time provenance binding is a defense for Model Context Protocol agent pipelines against cross-modal trust laundering, where instructions hidden in an email-signature image via LSB, DCT or metadata steganography get decoded by a tool. It combines HMAC-authenticated trust labels, a provenance ledger and context-taint tracking to block critical tool calls. On a real MCP testbed with 24 adaptive payloads, attack success fell from 100% undefended (62.5% for keyword and regex filters) to 0%.
The Decisions API is an OpenAI endpoint, built on GPT-6 Luna, that answers bounded questions about text, images or native JSON inputs, up to 10x faster than Luna through the Responses API. It returns a probability (predicate), a choice or a score. It is generally available, accepts up to 400 questions per request, and costs $0.10 per million input tokens with no output charges.
OpenAI APIDecisions APIGPT-Live+1
OpenAIGPT-6 Luna
llm-apps, agentsยท Published and added 22 Oct 2026
F2G-Pose is a direct, single-pass framework that predicts category-level object pose, metric size and camera-frame completed shape from one RGB-D image. It lifts frozen RADIO foundation features onto partial point clouds, tokenizes local geometry with DGCNN, and fuses it with a Mixture-of-Experts Transformer. It trains only on synthetic SOPE data, leads on SOPE and ROPE at 28 FPS, and is tested on HouseCat6D, HANDAL and real manipulation.
TransformerMixture of ExpertsDGCNN+2
F2G-PoseSOPEROPE+4
research, multimodalยท Added 24 Oct 2026, published 22 Oct 2026
bleepingcomputer.comarticle๐ก๏ธ Security
