LLMs & Generative AI - Page 30 of 64
Latest breakthroughs in large language models and generative AI shaping the future of artificial intelligence and machine learning.
1263 articles
View complete article list
OpenCode now supports Qwen3-Coder via config.json on Linux, macOS, Windows
OpenCode now quietly offers the one feature that matters: local AI. The Qwen3-Coder model runs on your own machine through Ollama, and a few lines in a config.json file will connect it to the editor.
Yelp expands AI chatbot to new Assistant tab across all categories
Yelp wants to be the first thing you open when you need something done in the real world. Its plan is a chatbot that does the booking, ordering, and scheduling for you. Today, that plan gets a lot bigger.
Qwen 3.6-35B-A3B Demo Implements Multimodal Inference, Thinking Control and RAG
Forget the binary of large and small. Qwen 3.6-35B-A3B redefines what’s possible by packing 35 billion parameters into a model that activates only 3 billion per forward pass. That’s efficiency without sacrifice.
Microsoft’s Phi-4-Mini 3.8B-Parameter Used in RAG Pipeline with LoRA Fine‑Tuning
Small models are rewriting the rules of production AI. Microsoft’s Phi‑4‑Mini packs just 3.8 billion parameters, yet it delivers reasoning, math, and tool‑calling chops that rival far larger architectures. But raw capability is only half the story.
OpenAI expands Trusted Access for Cyber Defense with GPT-5.4‑Cyber model
The line between offense and defense in cybersecurity has always been thin. OpenAI is now drawing a sharper one. With GPT-5.4‑Cyber, the company moves beyond general-purpose safeguards into a fine-tuned model purpose-built for verified defenders.
Moonshot AI, Tsinghua unveil PrfaaS KVCache that auto‑balances LLM nodes for throughput
Moonshot AI and Tsinghua have introduced PrfaaS, a cross-datacenter KVCache architecture that automatically rebalances prefill and decode nodes as traffic shifts.
Anthropic's Claude Opus 4.7 lifts coding benchmark 13% and solves four new tasks
Resolution jumps 13% on a 93-task coding benchmark. Four problems that stumped both Opus 4.6 and Sonnet 4.6 now fall. On CursorBench, the score clears 70%, a twelve-point leap from the previous generation.
OpenAI API guide demonstrates gpt-4o call, returning 'Late 2024-early 2025
The future of AI isn’t arriving in increments, it’s already here, and it answers with a timestamp.
Microsoft’s MarkItDown library converts zip files, unifying supported content
Every AI pipeline begins with a dirty secret: raw data is a mess. PDFs, Word docs, slides, images, audio files, spreadsheets , each demands its own conversion ritual before an LLM can digest them. Microsoft’s MarkItDown slices through that chaos.
NVIDIA KVPress Enables Long‑Context LLM Inference with KV Cache Compression
The appetite for long-context language models is insatiable, yet the memory they consume grows faster than our hardware can keep up. NVIDIA KVPress enters this fray not with a vague promise, but with a concrete mechanism to compress the KV cache.
Complete Real-World Example Shows Crawl4AI CSS Extraction and Filtering
Most web scrapers are junk. They fetch a blob of HTML and consider the job done, leaving you to hack through a jungle of irrelevant tags and scripts. The actual work starts after the download.
Implementing Context-Aware Long-Term Memory for AI Agents via Mem0 and OpenAI
Most AI agents today are brilliant, and completely forgetful. They treat every conversation as a blank slate, a clean room wiped clean the moment the chat ends. That’s not intelligence. That’s amnesia.
Transformer-Based Neural Quantum States for Frustrated Spins Using NetKet
Frustrated magnets are a computational dead end. The math is simple. Solving it for anything larger than a postage stamp of atoms isn't.
Schematik ‘Cursor for Hardware’ secures USD 4.6M Lightspeed; Anthropic wants in
Hardware has long been the slow, stubborn cousin of software, expensive to prototype, painful to iterate, and locked behind a wall of specialized knowledge. Schematik wants to shatter that wall.
Google AI launches Auto-Diagnose, LLM tool flags 84.3% of reports as ‘Please fix’
Auto-Diagnose is not gentle. Of 517 feedback reports, 436 came back with a blunt directive: “Please fix.” That’s 84.3%, an overwhelming majority. The tool isn’t polite. It’s precise. And developers seem to want it that way.
Penligent and Giskard among top AI red‑team tools for model security
Everyone's scrambling to lock down their AI. It's not magic; it's vulnerable code. Two names surface repeatedly in that panic: Penligent and Giskard. They lead a pack of nineteen tools, yes, but they embody the new, grubby reality.
AI protein-design tools offer flexible workflows for any protein class
Protein design tools used to be a scattered collection of niche instruments. That fragmented approach is finished.
Google's AI mode will open linked pages beside search on Chrome desktop
Google is now forcing external links from its AI mode to open in a side panel on Chrome desktop. Click one, and the page loads in a narrow window right beside the search dialog. The company detailed the shift in a blog post Wednesday.
Physical Intelligence robot model shows LLM-like skill composition, flaws noted
Robots are learning to improvise. Physical Intelligence’s latest model, π0.7, can recombine physical skills on the fly, much like a language model stitches together fragments of text.
OpenAI's Codex powers Lovable AI, letting millions create apps from text
The dream has always been simple: describe an app, and watch it build itself. For years, that was science fiction. Now, it’s a Tuesday afternoon. OpenAI’s Codex is the engine. Lovable AI is the vehicle.
Featured Resources & Reviews
No Code MBA Course Review
Learn to build AI-powered apps without coding. Our comprehensive review of No Code MBA's course.
AI Tools & Resources
Curated collection of AI tools, courses, and frameworks to accelerate your AI journey.
Weekly AI Digest
Get the week's most important AI news delivered to your inbox every week.