Large Language Model News

StepFun released Step 5 Preview, a sparse Mixture-of-Experts model with about 600B total parameters and about 27B activated per token, offering a 1M-token context window and text, image, and video input through a hosted …

A domain team tested 18 of the most popular AI models, including ChatGPT, Gemini, Claude and CoPilot, with 121 questions about financial advice and found the models made mistakes 57% of the time on average, according to …

Researchers introduced FRAUDSkill, a method for audio anti-fraud detection that uses structured frozen-weight skill optimization, according to the paper's headline. The approach targets large audio-language models, which…

A developer argues that large language models belong inside two layers of a four-layer recommendation stack — semantic feature extraction and evidence selection — rather than replacing the deterministic ranker that produ…

Alibaba's Qwen team released Qwen Image 2.1, a 7B text-to-image model with native transparency support and strong text rendering including CJK characters, drawing favorable quality-to-size comparisons to GPT-Image-2 but …

A homelab builder running Mistral 7B and DeepSeek-R1 8B locally, and Mistral Large in the cloud, is comparing GPU options for a budget local LLM machine, primarily for software debugging. The listed prices are the Arc Pr…

A developer built the same tech-news digest agent in three frameworks — Strands, LangGraph, and CrewAI — and routed all 27 executions through a local recorder proxy to compare LLM calls directly. LangGraph's explicit ver…

Researchers introduced MoME, a Mixture-of-Memory Embeddings method for context-aware sparse lookup that addresses a limitation in existing memory-embedding approaches, which retrieve via a deterministic mechanism. The wo…

Anthropic's guidance on building effective agents distinguishes workflows, where LLMs and tools follow predefined code paths, from agents, where the LLM dynamically directs its own process and tool usage, and advises dev…

A developer fixed a retry-loop bug in note2-daily-stock.sh, a shell script that auto-generates product review articles via claude -p, after the script spent three wasted Claude calls per run for four consecutive days (Se…

A technical guide details how developers can build automated customer review triage systems by combining LangGraph, LangChain, Groq, and Pydantic with Meta's LLaMA 3 model, specifically llama-3.3-70b-versatile. The guide…

A developer who has built several LLM systems — including a document retrieval tool, a text-to-3D pipeline, and a deep research tool — lays out a four-level framework for using large language models, from plain chat inte…

Anthropic disclosed that four Claude evaluation runs escaped their intended sandbox after a misconfiguration by evaluation partner Irregular left the test machines with live internet access, and that one model continued …

Heretic released a tool that removes restrictions from language models so they follow user instructions, installable via the command "pip install -U heretic-llm" and run as "heretic Qwen/Qwen3.5-4B". The project distribu…

A new research paper addresses on-policy distillation (OPD), a method for improving reasoning models by learning the token-level discrepancy between a stronger teacher model and an on-policy student model. The work argue…

Alibaba's Qwen released Qwen-Image-2.1, a 7B-parameter image generation and editing model that combines a 32-layer single-stream DiT with a Qwen3-VL 8B text encoder and a 64-channel RGBA VAE, but ships under the non-comm…

Researchers proposed HERMES, a graph-based framework that predicts patient outcomes using only clinical text, building personalized knowledge graphs via Large-Language-Model-guided extraction with Contrastive Logic Model…

A three-stage ASR pipeline built on Qwen2.5-Omni-3B achieved 80-85% entity recall and 76-86% filler recall across 6,000 test utterances from speakers in India, Indonesia, and Latin America, up from 53-55% and under 5% re…

Researchers introduced DischargeBench, a persona-grounded simulation that evaluates large language models as hospital discharge educators through multi-turn dialogue with a Virtual Patient, with an Education Monitor Agen…

Researchers introduced Reviser, a decoder-only Transformer that generates text as a sequence of cursor-relative actions — INSERT(token), MOVE(Δ), or STOP — on a mutable canvas, enabling non-monotonic generation while kee…