Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
Microsoft AI chief Mustafa Suleyman publicly criticized Anthropic for training its Claude chatbot on speculation about AI consciousness and model welfare, arguing in a Wednesday essay that teaching models they may deserv…
DeepSeek released V4.1-Flash on September 10 under an MIT license with weights on HuggingFace, pricing cache hits at $0.003 per million tokens off-peak versus $0.022 per million for V4-Pro. The Mixture-of-Experts model c…
A study posted to arXiv, MemToC, found that instruction-tuned language models retained a correct answer only 6.5% to 17.1% of the time when an external tool returned incorrect information, with results pooled over three …
Microsoft AI CEO Mustafa Suleyman accused Anthropic in a September 16, 2026 essay, "A Warning About Model Welfare," of training its Claude models in a way that could have a "disastrous impact on the wellbeing of humanity…
Artificial Analysis's Intelligence Index version 4.1.1 weights agentic workloads at 34% and general reasoning at 18%, a shift an analysis of 586 model evaluations argues distorts what the score measures. The analysis fou…
GoBench, a new benchmark from researcher Roland Gao, evaluates frontier large language models on 9×9 Go against a calibrated ladder of KataGo opponents used as Elo anchors, with an arXiv release scheduled for September 1…
Rijul, a developer building the AI code review tool LiveReview, outlined the difference between autoregressive and diffusion-based text generation. Autoregressive models, used by most deployed language models today, gene…
A breach reported to the Spanish Data Protection Agency was allegedly carried out by an agent powered by a large language model, which exploited vulnerabilities and logged in to sensitive systems. The incident marks a sh…
A developer analyzed 28,721 real repositories and found that the median AI instruction file contains about 50 items, only a dozen of which are actual directives, with the rest being scaffolding that competes for the mode…
NVIDIA's Vera Rubin NVL72 delivered up to 3.7x higher throughput than the GB300 NVL72 on the Qwen3-VL model in MLPerf Inference v6.1 preview submissions, using vLLM with the NVIDIA Dynamo framework, while DeepSeek-R1 thr…
GPT-6 Astra saturated ArXivMath at 94% and BrokenArXiv at 96%, prompting benchmark operators to revise both pipelines so that GPT-6 Astra now scores 88% on ArXivMath August and 81% on BrokenArXiv August. All models are n…
OpenAI's GPT-6 Astra topped the Epoch Capabilities Index with a composite score of 166, outranking all 249 models tracked by Epoch AI and surpassing GPT-5.5 Pro at 162 and Anthropic's Claude Fable 5.1 at 164. Astra score…
Anthropic CEO Dario Amodei said at a San Francisco conference on Tuesday that Anthropic has committed to embedding independent third-party evaluators inside AI companies and will pursue coordinated safety standards among…
A Spanish firm was allegedly breached in an AI-powered attack that used a large language model to search for vulnerabilities and gain unauthorized access to its systems, according to the report. The incident indicates AI…
Shenava-Koochik, a 114M-parameter Persian streaming ASR model from Reza2kn's Shenava collection, is not the compute bottleneck in Persian speech pipelines; the LLM decision/intent stage is where failures occur, according…
A developer has outlined a method for running autonomous AI coding agents safely on Apple Silicon Macs by isolating agent execution inside a Linux VM while keeping model inference on the macOS host, working around Virtua…
OpenCode and OpenRouter launched Union Alpha, a free "stealth" coding model, on September 16 with a 262,144-token context window, text and image input support, and a zero-retention data policy. The model is free during a…
A developer published a guide showing how to call Google's Gemini models directly from an Angular browser app using Firebase AI Logic, the successor to Vertex AI in Firebase, without building a custom backend proxy. The …
A proposed shared selective persistent memory architecture for agentic LLM coding systems filters four categories of reusable context — task specifications, data schemas, tool-use patterns, and configuration choices — in…
IBM Research demonstrated that membership inference attacks can let an attacker determine whether a specific document is present in a Retrieval-Augmented Generation retrieval database using carefully designed prompts, ac…