▲ Perplexity and NVIDIA just released JOLT, a post-training method where one model is both teacher and student, and the teacher is trained to teach. The key idea is simple. Self-distillation gives the model a verified answer and lets it score its own attempts. But a teacher holding the answer can take shortcuts the student cannot follow. JOLT trains the teacher with outcome reward plus a KL pull toward the student, so its feedback stays within reach. On MATH (Qwen3-0.6B), JOLT matched GRPO's accuracy with 6.5x fewer processed training tokens, and about 13x fewer completion tokens. JOLT+ (adds student reward) beat GRPO by 9.9 points on AppWorld and hit 10.96% pass@1 on Terminal-Bench 2.1 vs 8.25%. It is a trade-off though: without student reward, JOLT trails GRPO on AppWorld (51.92% vs 62.20%). It is a method, tested on Qwen3 0.6B to 30B-A3B and DeepSeek-V2-Lite. Full analysis: marktechpost.com/2026/10/21/… Paper: arxiv.org/abs/2610.10447 Blog: perplexity.ai/hub/blog/a-goo… @perplexity_ai @AravSrinivas

Datalab Introduces OmniExtractBench to Fix Bias and Opacity in Extraction Benchmarks 1 missed table row can drop an extraction score from 99% to 0%. Datalab just open-sourced OmniExtractBench to fix that. 620 documents, 4 vendor benchmarks, 1 deterministic scorer. Here's how it works. 👇 1. Pool the benchmarks 329 docs from LlamaIndex's ExtractBench, 47 from micro1, 42 from Extend's LongArray, 202 synthetic from Datalab. → Forms, filings, decks, 100+ page tables 2. Pair rows by content, not position Hungarian matching aligns table rows before anything is scored. → 100-row table, first row missed: 0% by position, 99% by content 3. 6 verdicts per value matched, misread, unfound, fabricated, invented_item, invented_field. → Every lost point traces back to 1 value 4. Nulls earn nothing Empty strings and None are dropped before scoring. → 100 padded empty fields add 0 verdicts 5. The numbers (10 configurations) → Datalab accurate: 93.85 → Datalab balanced 93.48 vs Reducto 93.47, effectively tied → GPT 5.6-sol: 95.11 precision, 84.99 recall Full analysis: marktechpost.com/2026/10/02/… GitHub Repo: pxllnk.co/hxplrq Technical details: datalab.to/blog/omni-extract… @datalabto @VikParuchuri

Google AI Releases Gemini 3.5 Transcribe: A Speech-to-Text Model Reporting 2.6% Average WER Across 85+ Languages They shipped two models instead of one. gemini-3.5-transcribe-live gives sub-second streaming over WebSockets but no diarization and no word-level timestamps, capped at 10-minute sessions. gemini-3.5-transcribe handles pre-recorded files with diarization, word offsets, and custom vocabulary up to 1,000 terms, at roughly half the cost. Key Takeaways: → Pick the endpoint before the model — latency and diarization are mutually exclusive here, and that shapes your whole pipeline → Smart mode can't be combined with timestamps or diarization — a readable summary and an auditable transcript are now two separate API calls → 4.0% WER streaming, 2.6% batch (per Artificial Analysis), 70% faster finalization than Chirp 3, 85+ languages, roughly $0.005–$0.009/min Full analysis: marktechpost.com/2026/08/27/… Technical details: marktechpost.com/2026/08/27/… @GoogleAI @googleaidevs

Vercel open-sourced the WebGPU library it built to ship the shaders on vercel. vgpu runs the same shader in the browser, in headless Node.js, and in your CI tests. • MIT licensed, on npm at v0.3.1 • .wgsl files import and export like TypeScript modules • Reflection generates bindings, no hand-written declarations • Headless Node.js rendering, Dawn-backed • Deterministic mock adapter for CI snapshot tests • 25 KB gzipped fullscreen effect, budget enforced in CI • CLI, agents.md, llms.txt, hosted read-only MCP endpoint • Examples include ONNX Runtime Web demos on WebGPU Full analysis: marktechpost.com/2026/08/28/… GitHub Repo: github.com/vercel-labs/vgpu Join our newsletter for similar product releases: aidevsignals.com/ @vercel @vercel_dev

The Video Production Stack Now Fits on One Desk: LTX-2.5 Launches as NVIDIA-Accelerated Open Weights World Model. It's an open weights world model for video, real-time apps, and physical AI — optimized with NVIDIA to run on RTX GPUs and DGX Spark. Here's what stood out: 1. The speed numbers are the story In LTX's published image-to-video benchmark (10-second clip): → 6.8 seconds on-prem (2x NVIDIA GB200) → 23.7 seconds via the LTX API → 52–70 seconds for the fastest closed rivals (Omni Flash, Grok 1.5, Veo 3.1) → 398 seconds for Kling 3.0 Pro — that's 58.5x slower On-prem generation finishes faster than the clip itself plays. 2. Multishot consistency fixes the real blocker Earlier open models generated each shot separately, so characters drifted between cuts — unusable for actual campaigns. LTX-2.5 renders the full sequence as one output, holding character, scene, and voice across cuts. A custom Gemma 4 backbone handles complex, multi-subject prompts. 3. Diffusion Fidelity Rendering is a smart cost tradeoff → Motion and structure built in an 8x temporally compressed latent space → Full detail spent only on high-fidelity keyframes → Keyframe count adapts to scene complexity Quality lands where it matters without full render cost on every frame. Full analysis: marktechpost.com/2026/08/11/… Model weight: huggingface.co/Lightricks Technical blog: blogs.nvidia.com/blog/local-… @NVIDIAAI @nvidia @NVIDIARobotics

How to Secure AI Agents, MCP Servers, and LLM Apps in Production Application security rests on one assumption: software does what its code says. ---AI agents broke it. Mend.io's new practitioner guide — 𝘚𝘦𝘤𝘶𝘳𝘪𝘯𝘨 𝘈𝘐 𝘢𝘨𝘦𝘯𝘵𝘴, 𝘔𝘊𝘗 𝘴𝘦𝘳𝘷𝘦𝘳𝘴 & 𝘓𝘓𝘔 𝘢𝘱𝘱𝘴 — starts from that break. An agent's behavior emerges from the model, the system prompt, retrieved context, and the tools it's permitted to call. The failure modes never appear in a CVE feed: prompt injection through data, over-permissioned agents causing damage without a single exploit, poisoned tool descriptions on MCP servers, EOL models serving predictions after patching stops. The guide's answer is three moves: 𝗦𝗲𝗲: Inventory the agentic attack surface across five layers — interaction, agent, integration, model, code. Hunt shadow agents via repo signatures and network egress. Run every agent through a 12-point misconfiguration checklist. 𝗙𝗶𝘅: Enrich → prioritize → triage. Rank by reachability and agentic amplification, not severity scores. Automate FP closures only with evidence trails. Risk acceptance is never automated. 𝗣𝗿𝗼𝘁𝗲𝗰𝘁: Guardrails on every input and output — embedded Python SDK or standalone Docker API server. Inbound: injection patterns, jailbreaks. Outbound: credentials, PII, policy violations. The core design principle: an agent that can't call a dangerous tool doesn't need a prompt begging it not to. Includes a 15-question maturity self-assessment aligned to NIST AI RMF, OWASP AIMA, ISO/IEC 42001, and the EU AI Act. Full analysis: marktechpost.com/2026/08/03/… Download the full guide, free: pxllnk.co/lxn88m @Mend_io

This World Model 'LingBot-World-Infinity (LingBot-World 2.0)' just released from Ant Group looks realy promising. It is an open causal world model with an Agentic harness. Most interactive world models hold together for a few minutes. Then textures smear and geometry warps. That's a video model, not a world — and Robbyant just drew the line at the attention mask. They released LingBot-World-Infinity (LingBot-World 2.0) — a 14B open causal video world model built on Wan2.2, trained with a Mixture of Bidirectional and Autoregressive (MoBA) attention mask, then distilled into a few-step real-time generator with no post-hoc drift filtering anywhere in the stack. Here's what's actually interesting: → Pure teacher forcing overfits — as context grows, the model leans on context instead of predicting frames. MoBA appends a bidirectional full-attention block as a regularizer → Leak-free cross-attention: AR rows attend to background prompt a_B plus chunk prompts a_≤i, lower-triangular. Bidirectional rows see one global prompt a_G → DMD runs over long self-rollout trajectories, not teacher-forced states — the student is optimized on the distribution its own errors induce → Director-Pilot harness: a VLM proposes event cards, the DiT generator renders physical dynamics. Mode B adds a SAM tracking loop for object-centric interaction → One 60-minute uninterrupted session, 20 distinct scenarios, no perceptible decay Full analysis: marktechpost.com/2026/07/09/… Paper: arxiv.org/pdf/2607.07534 Model weight: huggingface.co/robbyant/ling… GitHub Repo: github.com/robbyant/lingbot-… Project: technology.robbyant.com/ling… @AntGroup @robbyant_brain