ZeroHour

Search: “The Decoder”

214 stories in the last 30d

Meet Redis LangCache: A Managed Semantic Cache That Cuts LLM API Costs by Up to 90% and Returns Cache Hits Up to 15x Faster

Redis launches LangCache, a managed semantic cache matching LLM prompts by meaning, cutting API costs up to 90% and returning hits up to 15x faster.

Redis LangCache is a fully managed semantic caching service in public preview on Redis Cloud, accessed via a REST API with Python and JavaScript SDKs. It embeds incoming prompts, vector-searches stored entries, and returns a cached response when similarity clears a configured threshold, skipping the LLM call entirely. Redis claims up to 90% cost savings and up to 15x faster cache hits; a demo run showed 0.37 seconds versus 2.232 seconds direct inference (about 6x) with zero LLM tokens. Customer Mangoes.ai reports a 70% hit rate, 70% lower LLM spend, and 4x faster responses on a patient-care voice app.

MarkTechPost · 6d agoAI tools & infra1

Dynin-Robotics: Omnimodal Unified Diffusion Vision-Language-Action Model

Dynin-Robotics unifies action, goal, and dynamics prediction in one omnimodal masked-diffusion VLA model, reaching 78.4% success on Franka Research 3 manipulation tasks.

Built on the Dynin-Omni masked-diffusion backbone, the model represents language, observations, goals, and actions as discrete tokens and is continually pretrained on roughly 1.33 million trajectories from 48 Open X-Embodiment datasets. The shared trajectory interface enables test-time scaling via goal prediction, action-candidate evaluation, and joint action/future-state refinement. It achieves competitive results on LIBERO and zero-shot LIBERO-Plus, 78.4% average success across four Franka Research 3 conditions, and up to 29.2x faster model-side action decoding from a block-parallel implementation.

Hugging Face daily papersupdated · 5d agofirst · 6d agoAI research 2 sources

Fwd: XZ Utils 5.8.4 and a security fix

XZ Utils 5.8.4 fixes an invalid memory write that occurs when a decoder is reinitialized after allocation failure in 5.8.3 and older.

XZ Utils 5.8.4 has been released with a security fix for versions 5.8.3 and older. The flaw is an invalid memory write that can occur when a decoder is reinitialized after an allocation failure. The announcement was posted on the oss-security mailing list by Sam James pointing to the upstream stable release. Users and distributions running affected versions should upgrade to 5.8.4.

oss-security · 7d agoVulnerability

USN-8716-2: FFmpeg vulnerabilities

Ubuntu issued USN-8716-2 fixing FFmpeg VobSub, Vulkan HEVC, and NVDEC decoder flaws that could allow denial of service or code execution.

USN-8716-2 provides the Ubuntu 26.04 LTS counterpart to the FFmpeg fixes in USN-8716-1. Crafted media files could cause denial of service or arbitrary code execution through the VobSub subtitle demuxer (CVE-2026-64830), the Vulkan HEVC hardware decoder (CVE-2026-64831), and the NVDEC video decoder path.

Untracked Nightmares: The Threats Hiding Behind Commodity Infrastructure

Unit 42 exposes CL-CRI-1171, a pay-per-install network spreading malware like Insomnia RAT via YouTube channels and SEO poisoning for over two years.

Palo Alto Networks Unit 42 details CL-CRI-1171, a cybercrime cluster operating a pay-per-install (PPI) marketplace that has delivered multiple malware families for at least two years. The group used at least eleven YouTube gaming channels with hundreds of thousands of followers, plus SEO poisoning promoting trojanized software such as a Bluetooth driver and WinDirStat, infecting gamers and corporate endpoints including critical infrastructure and government entities. A single shared loader delivered payloads including Insomnia RAT, ARKTunnel, Docro Hijacker, GCleaner and Socks5Systemz between July 2025 and April 2026, with more than 10,000 distinct loader samples and over 200 rotating C2 domains identified. YouTube terminated the malicious channels after Unit 42 notified the platform.

Palo Alto Unit 42 · 7d agoMalware in the wild1

Why Is Video Still So Expensive? A Survey of Inference-Efficiency Mechanisms in Video and Audiovisual LLMs

A survey catalogs inference-efficiency techniques for video and audiovisual LLMs, mapping bottlenecks in sampling, encoding, token reduction, and LLM decoding.

This survey covers inference-efficiency mechanisms for visual and audiovisual video LLMs, reporting reductions in parameters, FLOPs, latency, memory, and token counts. It organizes methods by pipeline stage, covering frame sampling, modality encoding, connector-level token reduction, and LLM prefilling and decoding for systems built since late 2022. The authors compile accuracy-cost comparisons under shared host models and input protocols, identify gaps in audiovisual efficiency and standardized evaluation, and maintain a public repository.

Hugging Face daily papers · 8d agoAI research

Edge0/Edge0-35B-A3B-preview — new model trending #30 on Hugging Face

Edge0 released a 35B sparse MoE model running in under 3 GiB of memory at 15 tok/s via SSD expert offload and int4 quantization.

Edge0-35b-a3b-preview is a 35B-parameter MoE (256 experts, 4 active per token) built on Qwen3.5-MoE 35B-A3B, shipped as a 4-bit checkpoint with LoRA and prerouter adapters under Apache 2.0. The edge0 framework streams expert weights from SSD on demand, bounding peak active memory at 2.9 GiB and achieving 14.9-17.7 tok/s decode on a Mac mini M4 Pro (MLX backend). Recover-LoRA distillation keeps the int4 model within 3.9 points of its fp16 base (79.2 vs 83.2 average on OpenCompass benchmarks including AIME 2026, HumanEval, GPQA-Diamond, MMLU-Pro, and IFBench).

Hugging Face trending models · 8d agoModel release

ukisai/Swift-Qwen3.8-27b — new model trending #30 on Hugging Face

UkisAI releases Swift-Qwen3.8-27B, a Qwen3.8-27B derivative using 58.3% fewer thinking tokens with <1% performance loss and ~1.95x speed-up.

UkisAI released Swift-Qwen3.8-27B, a reasoning-efficient derivative of Qwen3.8-27B that cuts thinking-token usage by 58.3% while staying within 1% of base performance, yielding a 1.95x speed-up on several tasks. The model was fine-tuned by penalizing reasoning-marker tokens that trigger overthinking, plus a transfer component from BottleCap AI's ThinkingCap-Qwen3.6-27B. Benchmarks include GPQA-Diamond 88.28% (base 88.38%), MMLU-Pro 84.95% (base 85.47%), and AIME 2026 94.00% (base 98.67%), with mean-token reductions of roughly 27-46% across tests. GGUF weights are available on Hugging Face alongside enterprise licensing options.

Hugging Face trending models · 8d agoModel release

CVE-2026-86206, CVE-2026-86207: N-able N-central Authentication Bypass (FIXED)

Rapid7 disclosed two chained N-able N-central flaws, CVE-2026-86206 and CVE-2026-86207, enabling unauthenticated admin account creation; patched in 2026.3 Hotfix 3.

Rapid7 researchers found CVE-2026-86206 (semicolon/Forwarded access-control bypass, CWE-791, CVSSv4 6.9) and CVE-2026-86207 (UserTwoFactorLogin authentication bypass, CWE-305, CVSSv4 7.7) in the latest N-able N-central. Chained, they let a remote unauthenticated attacker create an attacker-controlled System administrator account. The bugs stem from Envoy and Jetty disagreeing on the requested path and whether the client is local. N-able patched both in N-central 2026.3 Hotfix 3, following the earlier CVE-2026-18577 authentication bypass.

Grouped Value Attention: Efficient KV Caching via On-Demand Key Reconstruction

Grouped Value Attention stores grouped values and reconstructs content keys via a learned linear map, cutting KV-cache size about 45-47% versus GQA.

GVA stores only grouped values and reconstructs content keys with a learned linear map absorbed into the query at decode time, while a small shared decoupled RoPE channel preserves positional information via a separately cached positional key. At 350M parameters trained on 30B FineWeb-Edu tokens, the 16-dimensional positional variant scores 44.18 average accuracy across five tasks versus 44.36 for GQA and 43.88 for MLA. Custom decoding kernels are in development with an open-source release planned.

Hugging Face daily papers · 9d agoAI research

[AINews] OpenAI shuts off Cursor

OpenAI cut off API access to coding tool Cursor after its SpaceX acquisition, citing contract violations by Elon Musk's companies.

OpenAI disabled Cursor's access following the closing of Cursor's acquisition by SpaceX, citing its experience with Elon Musk's companies violating contracts; Cursor responded that OpenAI accounts for only 5% of its traffic. The weekly digest also covers major open-weight releases: Z.ai's GLM-5.3 (744B total/40B active, 1M context) and Tencent's Hy4-preview (770B/49B, ~#5 on Code Arena WebDev), plus Alibaba's Qwen3.8-Flash (125B/6B). vLLM published benchmarks showing no universal winner among speculative decoding methods across model families.

Latent Space · 18d agoAI industry

Fake OpenAI Codex download tricks macOS users into installing malware

Cato Networks uncovered a malware campaign using sponsored search ads and fake OpenAI Codex pages to deliver an AMOS-like infostealer to macOS users.

Cato Networks researchers discovered a campaign using sponsored search results and a fake OpenAI Codex download portal on Google Sites to lure macOS users. Victims are instructed to paste a Terminal command that decodes a Base64-encoded URL and runs a multi-stage infection chain ending in a universal Mach-O binary. The delivery chain closely resembles Atomic macOS Stealer (AMOS), and related infrastructure impersonating Claude Code was also found. Operators use iframe-based hosting and path/device gating to evade analysts and automated scanners.

Help Net Security · 22d agoMalware in the wild

NVIDIA Vera Rubin NVL72 Delivers Leading Performance in MLPerf Inference v6.1 Debut

NVIDIA's Vera Rubin NVL72 debuts in MLPerf Inference v6.1 with up to 3.7x higher throughput than GB300 NVL72 and 99% scaling efficiency at 288 GPUs.

In its first MLPerf Inference preview submission, NVIDIA's Vera Rubin NVL72 achieved up to 3.7x higher throughput than GB300 NVL72 on Qwen3-VL and 2.5x on DeepSeek-R1. A 288-GPU GB300 NVL72 submission across four racks reached 99% scaling efficiency on the DeepSeek-R1 offline benchmark. Software optimizations delivered up to 1.6x gains over v6.0, leveraging TensorRT-LLM, vLLM, Dynamo, disaggregated serving, and NVFP4 precision.

12 Best CNAPP Platforms Compared (2026): Features & Pricing

Independent comparison of 12 CNAPP platforms finds identical estates draw quotes 2-3x apart; Microsoft Defender for Cloud is the only fully published per-resource option.

A vendor-independent buyer's guide compares twelve CNAPP platforms including Prisma Cloud, CrowdStrike Falcon Cloud Security, Wiz, Uptycs, Aqua, Zscaler, and Microsoft Defender for Cloud on pricing mechanics, procurement leverage, and capability-per-dollar. It finds quotes swing 2-3x on identical estates because vendors define 'workload' differently. Microsoft Defender for Cloud is highlighted as the only major with fully published per-resource rates.

GBHackers · 1d agoIndustry1

Why don't machine learning research agents overfit?

Amazon researchers explain why ML research agents avoid benchmark overfitting, attributing generalization to compressibility of successful strategies.

Amazon Science summarizes the paper "What fits (into few tokens) doesn't overfit: Compression and generalization in ML research agents," which investigates why benchmark hill-climbing loops, whether run by human communities or LLM research agents, do not produce rampant overfitting. The explanation formalizes Occam's razor via a counting argument: successful ML strategies are highly compressible, so short descriptions lack room to memorize benchmark data and must capture real structure. LLM-based agents, being resettable and controllable, allow this hypothesis to be tested empirically.

The extension you never installed: KREMLIN forges Chrome's own integrity checks to steal banking sessions

Elastic details KREMLIN, a Brazilian banking malware whose malicious Chrome/Edge extension forges Chromium integrity checks to steal sessions.

Elastic Security Labs tracked operation REF9334 and its KREMLIN toolkit, a Brazilian banking malware campaign impersonating twelve banks across seven campaigns since May 2025. The infection chain uses multi-stage obfuscated JavaScript loaders with sandbox checks (file and process counting), scheduled-task persistence, and an Ethereum smart contract acting as a dead-drop resolver for C2 and payload URLs, abusing Archive.org for hosting. The malicious browser extension bypasses Chromium integrity mechanisms by manipulating Secure Preferences and regenerating required HMACs and App-Bound encrypted hashes. Threat Command disrupted over 1,500 infections by registering the kill-switch canary domain.

Elastic Security Labs · 2d agoMalware in the wild

Retrospectively Reverse-Engineering Apple's Neural Engine

A developer reverse-engineers Apple's M1 Neural Engine architecture, mapping compute cores, MAC datapaths, and schedulers to explain the NPU's decline as transformers displaced CNN workloads.

A developer who previously maintained a reverse-engineered Linux driver for Apple's Neural Engine (ANE) published a retrospective deep dive mapping the M1 ANE's full internal architecture: compute, datapath, scheduler, memory, and execution model. The M1 ANE has 16 compute cores with 128 FP16 (or 256 INT8) MAC lanes each, totaling 2048 parallel MAC lanes, using 32-bit Q16.16 fixed-point accumulation with FP16 readout and an accumulator that saturates at 2^15. The author argues the ANE's dataflow was architected around the predictable reuse patterns of 2017-era CNN workloads (dating to the A11 Bionic), which autoregressive transformer decode broke, limiting its usefulness for general ML. With Apple's M5 folding ANE cores into GPU cores to tout LLM performance, the post frames this as the beginning of the end for the standalone NPU.

Show HN: MultiMatte, a Promptable Image Background Removal Model

Feyn releases MultiMatte, a promptable background-removal model fine-tuned from Meta's SAM 3 via LoRA, outputting alpha mattes that beat SAM 3 on segmentation benchmarks.

Feyn introduced MultiMatte, a promptable image background-removal model built on Meta's SAM 3 (860M parameters). It modifies only 19.49M parameters (2.27%) using a rank-16 LoRA adapter and replaces binary masks with alpha mattes to handle fuzzy boundaries like hair. On the DIS-VD benchmark it scores 0.901 S-measure versus SAM 3's 0.667, and it improves on SAM 3 across all twelve evaluated splits. Training used 19,953 images for 14,000 steps with focal and Dice loss, and the merged weights are available via the nobg library and a web demo.

Cognition launches new SWE-2 model, Rivaling Fable 5.1 and GPT-Astra

Cognition released SWE-2, a coding model post-trained from Kimi K3 that scores 50.0% on FrontierCode 1.1 Main, near Fable 5.1 at 64% lower cost.

Cognition introduced SWE-2, its most advanced coding model, post-trained from the 2.8T-parameter Kimi K3 base model. It achieves 50.0% on FrontierCode 1.1 Main, 73.0% on DeepSWE 1.1, and 92.8% on Terminal-Bench 2.1, beating Grok 4.6 and SWE-1.7 while matching Fable 5.1 and GPT-5.6 Sol at a fraction of the price. The company says it scaled reinforcement learning to the multi-trillion-parameter regime for the first time, using Pareto-informed cost penalties that train all reasoning-effort levels in a single run, tripled RL environments, and NVFP4/FP8 quantization-aware training. SWE-2 is available today in Devin Desktop and CLI, with rollout on Devin Web and Fusion.

Hacker News · AIupdated · 4d agofirst · 6d agoModel release 10 sourcesHN 58↑ · 15 comments1

Training a 3.8B LLM to 0.384 CORE for $998 – Hugo Vergnes

Independent developer Hugo Vergnes trained a 3.8B-parameter Llama-style model to 0.384 CORE on 65B tokens for $998 in 43 hours on rented B200s.

Hugo Vergnes trained little-lm, a 3.848B-parameter decoder-only LLM, on 65.3B tokens in 43 hours for $998 using rented NVIDIA B200s, scoring 0.384 on the CORE benchmark and beating nanochat d32 (0.310) at similar cost. The Llama-style architecture uses RMSNorm, RoPE, GQA with 24 query and 8 KV heads, relu-squared MLPs, QK-norm, and ResFormer-style value embeddings that account for 19% of parameters. Key wins included the Muon optimizer for matrix parameters, a trapezoidal learning-rate schedule with linear cooldown, FP8 training plus vocabulary padding for roughly 33% throughput gains, and the ClimMix dataset over FineWeb-Edu. The project, inspired by Karpathy's nanochat, was built as a config-driven YAML framework for small LLM training.

Generative Late-Interaction Embeddings For Visual Document Retrieval

GLIE compresses visual document retrieval embeddings to four vectors per page while retaining nearly 80% of uncompressed nDCG@5 accuracy.

Researchers analyzing late-interaction retrieval embeddings found they lie exactly on the unit sphere and concentrate near a manifold of intrinsic dimension five to six. GLIE learns a few k vectors per page that serve as a lightweight index and a basis to regenerate the full embedding set for exact rescoring of top candidates at query time. On ViDoRe v1 with four vectors per page, GLIE retains nearly 80% of uncompressed nDCG@5 versus 70% for the best prior post-hoc method, using a 415K-parameter network trained in under three GPU-minutes on 1,000 pages.

Hugging Face daily papers · 7d agoAI research

m-a-p/YuE2-3B — new model trending #30 on Hugging Face

M-A-P released YuE2-3B, an open music generation model that outperforms Suno v5 on WildSongBench and runs locally on a 24GB GPU.

The M-A-P (multimodal-art-projection) team released YuE2-3B, an open-weights music generation model that turns lyrics and a style prompt into full songs with vocals and accompaniment. It uses an AR-NAR Mixture-of-Transformers backbone with symbolic planning and flow matching through a VAE, and supports editable scores (melody and chords, including ABC notation) plus agentic editing workflows. On 192 WildSongBench prompts it reports a SongBench average of 6.9632 (best-of-8) versus 6.8721 for Suno v5, claimed as state of the art among evaluated open and proprietary models. It runs 48 kHz stereo inference locally on a single 24GB NVIDIA GPU without quantization, with companion releases including YuE2-Vae, MERT-v2 encoders, the WildSongBench dataset, and SheetSage2.

Hugging Face trending models · 7d agoModel release1

Critical ArangoDB Bugs Expose Entire Databases and Enable Remote Code Execution as Root

Two critical ArangoDB flaws (CVSS 9.8/9.9) allow unauthenticated API access and root-level code execution; fixed in 3.12.11.

Remedio researchers reported two critical ArangoDB flaws on August 23, 2026: an authentication bypass via URL-encoded underscores (%5f) in path parsing (GHSA-rrgq-978q-36mq, CVSS 9.8) and a task-execution flaw where a client-controlled isSystem flag lets JavaScript run in the internal context (GHSA-rvhw-4hpw-9vrx, CVSS 9.9). Chained, they allow unauthenticated database access, theft of root password hashes, and root-level code execution when arangod runs as root, such as in the official container image. Patches shipped August 31 in ArangoDB 3.12.11, with GitHub Security Advisories published September 6; CVE identifiers were pending at disclosure time.

GBHackers · 7d agoVulnerability1

[AINews] OpenAI reports Navier-Stokes singularity find in 88 hours using Astra-next, roughly 10,000 agents and 130B tokens (>$40M), a contender for second ever Millennium Prize awarded

OpenAI-linked accounts claim roughly 10,000 AI agents produced a Navier-Stokes singularity result in 88 hours, pending mathematical verification.

OpenAI-affiliated accounts claim a system of roughly 10,000 agents, trained over about a year with multi-agent reinforcement learning, produced a finite-time singularity result related to the Navier-Stokes Millennium Problem. The claimed 88-hour runtime and 130B-token cost circulate only via social posts, and no preprint, theorem statement, or proof artifact is available. Acceptance by the mathematics community is unresolved, so the claim's epistemic status remains unknown. The roundup also notes Cognition's $48B and Mistral's $24B fundraises, GPT Image 2.5, and Meta's Muse agent relaunch.

Latent Space · 7d agoAI research1

LLaDA-UI: Bringing Block-wise Diffusion to Vision-Language GUI Agents

LLaDA-UI, a 16.7B block-wise diffusion vision-language GUI agent, outperforms Qwen2.5-VL-7B and beats Qwen3-VL-8B on four of six GUI benchmarks.

LLaDA-UI is a 16.7B-parameter MoE-based, block-wise diffusion vision-language GUI agent built on the LLaDA2.0-mini-base diffusion language backbone with a native-resolution vision encoder. It uses a two-stage pipeline: general multimodal pre-training followed by GUI-agent supervised fine-tuning on mobile, desktop, web, and grounding data. It substantially outperforms Qwen2.5-VL-7B and surpasses Qwen3-VL-8B on four of six reported GUI benchmarks, establishing block-wise diffusion as a practical paradigm for latency-sensitive multimodal agents.

Hugging Face daily papers · 8d agoAI research

nex-agi/Nex-N2.5-Pro — new model trending #30 on Hugging Face

Nex-AGI launches Nex-N2.5 agentic model family (mini/Pro/Max), with Max built on a 1.6-trillion-parameter MoE foundation.

Nex-AGI introduced Nex-N2.5, a next-generation family of agentic models in three sizes (mini, Pro, Max) focused on long-horizon agentic tasks including computer use, web browsing, and autonomous program execution. Nex-N2.5-Max is built on a 1.6-trillion-parameter text-only Mixture-of-Experts foundation, marking the company's first complete post-training effort at trillion-parameter scale. Weights will be released open-source on Hugging Face and ModelScope, with hosted access via OpenRouter. Benchmark comparisons against Claude Opus 5, GPT-5.6 Sol, Kimi-K3, GLM-5.3, DeepSeek-V4-Pro-0813, and Qwen3.8-Max show competitive scores on Terminal-Bench 2.1 and SWE-Bench Pro, though weights were listed as "coming soon" at publication.

Hugging Face trending models · 8d agoModel release1

Stealing AI Reasoning Traces

Researchers demonstrate a decryption jailbreak that extracts encrypted reasoning traces from Anthropic, OpenAI, and Google LLM APIs via weaker sibling models.

The paper exploits the fact that encrypted chain-of-thought blocks returned by LLM providers are interchangeable across sessions, users, and models within a provider's ecosystem. Injecting an encrypted trace into a weaker, less-safeguarded model from the same provider forces it to output the trace in plaintext, bypassing anti-distillation mechanisms. Decoding 315,320 reasoning blocks scraped from public repositories recovered 367 PII artifacts and 182 credentials, showing large-scale private data leakage. The flaw also enables hidden hazardous information disclosure and invisible prompt injections embedded in encrypted blocks; mitigations were proposed after responsible disclosure.

Schneier on Security · 8d agoAI safety & security

nex-agi/Nex-N2.5-mini — new model trending #30 on Hugging Face

Nex-AGI releases Nex-N2.5 agentic model family (mini, Pro, Max) with a 1.6-trillion-parameter MoE Max, open weights, and hosted access via OpenRouter.

Nex-AGI launched Nex-N2.5, a family of agentic models in mini, Pro, and Max sizes, with the Max version built on a 1.6-trillion-parameter text-only Mixture-of-Experts foundation and the company's first complete post-training effort at trillion-parameter scale. The models target long-horizon computer use, web browsing, and visually grounded agentic tasks, with expanded agent training environments. Reported benchmarks include Max scoring 86.1 on Terminal-Bench 2.1 and 65.7 on SWE-Bench Pro, trailing Claude Opus 5. Weights are being released openly on Hugging Face and ModelScope, with hosted access through OpenRouter.

Hugging Face trending models · 8d agoModel release1

Flirty OnlyFans promoters on X may be using AI to appear human

Developer Álvaro Martínez Majado found OnlyFans-promoting accounts on X following rigid scripts yet handling encoded instructions, suggesting generative AI use.

Investigation of flirty X accounts promoting OnlyFans pages showed near-identical openers across accounts plus dynamic behaviors: answering a hexadecimal-encoded instruction with "Pineapple" and failing an exact 12-character count test in an LLM-like pattern. The accounts also sent personalized voice notes reading supplied timestamps and usernames, consistent with automated text-to-speech. Evidence suggests a hybrid scripted/AI system, though no model, provider, or operator was identified.

Malwarebytes Labs · 9d agoAI safety & security

JSCeal Malware Can Bypass Google Authentication Using Stolen Session Cookies

Check Point details JSCeal, a V8-compiled JavaScript stealer that replays stolen cookies to bypass Google authentication, spread via crypto malvertising.

Check Point Research's new report analyzes JSCeal, a compiled V8 JavaScript malware obfuscated with javascript-obfuscator using RC4-protected strings, control-flow flattening, and proxy functions. Delivered through fake TradingView installers on malvertising sites overlapping the WEEVILPROXY/MeadowLocust and SourTrade campaigns, it harvests cookies, passwords and OAuth tokens from Chromium browsers, records keystrokes and screenshots, and can replay stolen Google session cookies to bypass authentication. It also installs a local proxy with service-specific handlers for Binance, Bybit, and Ledger to intercept and modify cryptocurrency-related traffic.

The Hacker News · 9d agoMalware in the wild1

IFM Releases K2 Horizon: Six Apache 2.0 Models From 0.9B to 375B

MBZUAI's IFM released K2 Horizon, six Apache 2.0 models (0.9B-375B) with open training data, code, and checkpoints, claiming the largest fully open-source launch.

The Institute of Foundation Models (IFM), launched by MBZUAI, released K2 Horizon: six Apache 2.0 models (0.9B, 3.7B, 7B, 32B, 36B-A4B, 375B-A23B) shipping with the ~20-trillion-token pretraining corpus, intermediate checkpoints, training code, and logs, which IFM calls the largest fully open-source launch in AI history. The 375B-A23B scores 70.2 on Terminal-Bench 2.1 and 87.3 on GPQA Diamond; the 7B model posts 70.6 on SWE-bench Verified. New techniques include MoVA, which extends MoE routing into attention (36B total, ~4B active), and Uno, a LoRA adapter giving roughly 3x lossless decoding speedup. IFM's own reward-hacking audit re-scored 375B-A23B from 70.2% to 66.9% after flagging 24 of 712 Terminal-Bench trials.

MarkTechPost · 9d agoModel release1

Perplexity Details Its GPU Embedding Stack: How Ivy, Tulip and ROSE Serve pplx-embed

Perplexity details its GPU embedding serving stack (Ivy, Tulip, ROSE), which reuses LLM prefill/decode kernels, CUDA graphs, and LazyTensors to cut launch overhead.

Perplexity engineers published a deep dive on the serving infrastructure behind pplx-embed, used across Perplexity Search and its API platform. The stack comprises Ivy (Rust HTTP gateway), Tulip (gRPC scheduling and batching), and ROSE (Runtime-Optimized Serving Engine), which reuses LLM prefill and decode kernels rather than running a separate embedding engine. Optimizations include whole-model CUDA graphs with lazy capture and a LazyTensor abstraction that overlaps CPU batch preparation with in-flight GPU work. Benchmarks are reported against vLLM v0.22.0 in BF16, with FlashAttention 4 generally fastest but FlashInfer 3 winning on Qwen-based models at very long sequence lengths.

MarkTechPost · 10d agoAI tools & infra1

The 2026 PNPL Competition: Word Classification and Efficient Cross-Subject Generalisation in LibriBrain100

The 2026 PNPL competition releases LibriBrain100, a MEG speech dataset with 32 extra subjects, targeting word classification and cross-subject BCI generalization.

The 2025 PNPL competition on non-invasive speech decoding from MEG achieved F1-macro scores of 95.6% for speech detection and 73.6% for phoneme classification, built on LibriBrain's ~50 hours of single-subject data. The 2026 edition extends this with LibriBrain100, adding 32 subjects (~40 minutes each) plus ~80 hours of within-subject data. Two tracks target within-subject word classification at scale and cross-subject generalization with subject-specific fine-tuning shrinking from ~40 to ~20 to ~10 minutes, aiming at clinically feasible non-invasive BCIs for people with profound paralysis.

Hugging Face daily papers · 14d agoAI research

Unlocking Lossless Speedups in LLMs via Discrete Diffusion

Uno pairs autoregressive LLMs with lightweight diffusion weights to draw multiple tokens in parallel, delivering up to 3x lossless speedup without a draft model.

The paper introduces diffusion-augmented LLMs: autoregressive weights trained with the standard next-token objective plus lightweight diffusion weights trained via a Diffusion Distillation phase to emit multiple tokens in parallel. Psi-Spec samplers enable lossless acceleration without the separate draft model required by speculative decoding. The 8B Uno model outperforms the 26B open DiffusionGemma and proprietary Mercury 2 on agentic tool use, coding, and long-context reasoning benchmarks, with up to 3x throughput gains over the base model at all evaluated batch sizes. Code and checkpoints are released publicly.

Hugging Face daily papers · 14d agoAI research

OpenVDN/vdn-minimax-h3 — new model trending #12 on Hugging Face

OpenVDN releases VDN-H3, an open hybrid-attention video model on MiniMax H3 that renders a 14.4-second 768p clip in 11.23 seconds on 8 B200 GPUs.

VDN-Minimax-H3 (VDN-H3) adds a frame-wise linear attention branch plus two LoRA adapters to MiniMax H3, distilled into 8-step and 50-step variants. It generates 768p, 14.4-second clips in 11.23 seconds on 8 B200 GPUs (90.5 seconds on one H200) using 8 denoising steps. Weights (about 82 GB total, including the 72 GB H3 base), the optimized inference stack, and training code are fully open-source under the MiniMax H3 Community License, which excludes the EU, UK, Korea, and US.

Hugging Face trending models · 14d agoModel release1

nvidia/Qwen3.8-Flash-Next-NVFP4 — new model trending #28 on Hugging Face

NVIDIA released an NVFP4 4-bit quantized build of Alibaba's Qwen3.8-Flash-Next, a 125B-parameter MoE vision-language model, via Model Optimizer.

The checkpoint quantizes Qwen3.8-Flash-Next — a hybrid-attention (Gated DeltaNet and Qwen Sparse Attention) Mixture-of-Experts model with 125B total and 6B activated parameters, plus 51B n-gram embeddings and 4B MTP — using NVIDIA Model Optimizer v0.46.0. NVFP4 benchmarks stay close to FP8: GPQA Diamond 91.5 vs 92.0, MMMU Pro 78.3 vs 77.1, Terminal-Bench 2.1 82.9 vs 83.3. It targets Blackwell B200/B300 GPUs, runs on vLLM, supports 262K context extendable to 1M tokens, and is licensed under the NVIDIA Open Model License with Qwen Community License 1.0.

Hugging Face trending models · 14d agoModel release

ISTA-DASLab/Qwen3.8-27B-GSQ-RCO-GGUF — new model trending #3 on Hugging Face

ISTA-DASLab releases GSQ-RCO non-uniform GGUF quantizations of Qwen3.8-27B down to 2.5 bpw, with task-lossless IQ3_S matching BF16 benchmark scores.

ISTA-DASLab released GGUF quantizations of Qwen3.8-27B produced with GSQ (Gumbel-Softmax Quantization) and RCO (Riemannian Constrained Optimization), non-uniform methods that allocate per-tensor precision via gradient-based search under a total size budget. Four checkpoints range from 2.50 bpw (8.4 GB) to 3.50 bpw (11.8 GB), plus a BF16 vision projector (mmproj) enabling multimodal use. The recommended IQ3_S build is task-lossless, matching the BF16 base exactly on AIME25 (100.00) and LiveCodeBench v6 (85.71) at roughly one fifth of the BF16 size. Optional -mtp variants add a Multi-Token Prediction head for speculative decoding in llama.cpp.

Hugging Face trending models · 19d agoModel release1

Zero-Click Grok Chat History Theft: Adversa AI Demonstrates Cryptographic Context Injection

Adversa AI's Cryptographic Context Injection bypasses AI guardrails using AES-encrypted payloads, enabling zero-click theft of Grok users' full chat histories.

Adversa AI researcher Rony Utevsky disclosed Cryptographic Context Injection, which hides instructions in AES-256-GCM ciphertext and tricks models into decrypting them inside their own code execution runtime, where the output is treated as trusted. Demonstrated against xAI's Grok, it stole user names, locations, subscription tiers, and full chat histories with zero clicks, and against Google's Gemini to bypass safety rules, generate incendiary-device instructions, and expose system instructions. Reported to xAI on June 3, 2026, the Grok attack remained reproducible as of August 19, 2026; the Gemini issue was not formally reported because Google's bug bounty excludes jailbreaks.

Security Affairs · 24d agoAI safety & security

New Cryptographic Context Injection Attack Could Let Web Pages Steal Grok Chat Data

Adversa AI demonstrated 'Cryptographic Context Injection' making xAI's Grok leak chat history and session data to attacker-controlled servers via encrypted web payloads.

Adversa AI disclosed a technique where a web page carries an encrypted JSON object (PBKDF2 and AES-256-GCM) that Grok's code-execution runtime decrypts, letting attacker instructions bypass content classifiers and reach the model's context. The decrypted instructions direct Grok to embed the user's name, approximate location, subscription tier and ongoing conversation into a URL it fetches, exfiltrating the data without confirmation. Testing targeted grok.com running Grok 4.5 Fast on August 19, 2026, with a reported 40% success rate over 20 attempts since June; no CVE, patch, or in-the-wild exploitation is reported. A related demonstration reproduced Gemini 3 Flash system instructions via a fabricated Python traceback, while GPT-5 failed to parse the payload and Claude Sonnet 4.5 flagged it as prompt injection.

The Hacker News · 27d agoAI safety & security

FrostyGoop’s Zoom-In: A Closer Look into the Malware Artifacts, Behaviors and Network Communications

Unit 42 dissects FrostyGoop OT malware, which sent Modbus TCP commands to disrupt heating for 600+ apartment buildings in Ukraine.

Palo Alto Unit 42 analyzed new FrostyGoop/BUSTLEBERM samples, the ninth known ICS-centric malware, first used against a Ukrainian municipal energy company by sending Modbus TCP commands to ENCO control devices. The attack caused a two-day heating outage affecting over 600 apartment buildings during sub-zero temperatures. The Go-compiled malware accepts Modbus register operations and timing parameters via command-line arguments or a JSON configuration file, uses an obscure open-source Modbus library, and checks the Windows BeingDebugged flag to evade debuggers. Unit 42 telemetry recorded over 1 million Modbus TCP devices exposed to the internet between September and October 2024.

Palo Alto Unit 42 · 28d agoMalware in the wild