Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
Researchers at the Association for Computational Linguistics introduced a Dual-Phase Self-Evolution (DPSE) framework for large language models that jointly optimizes user preference adaptation and domain-specific compete…
Researchers at the 64th Annual Meeting of the Association for Computational Linguistics introduced a dual-task paradigm combining arithmetic and sentence comprehension to study how language models allocate cognitive reso…
Researchers introduced DySECT, a dynamic self-evolving extraction system that uses a closed-loop cycle where an LLM extracts triples to populate a knowledge base, which then improves the LLM via prompt tuning or fine-tun…
Researchers at ACL 2026 found that entropy collapse in LLM reasoning, which undermines test-time scaling, is driven by premature overconfidence at a small set of critical tokens. They proposed SCOPE, a method that applie…
Zhipu AI's open-weight GLM 5.2 model scored 39% F1 on IDOR detection in Semgrep's benchmarks, outperforming Claude Code (32%) at $0.17 per vulnerability. The model, released June 13, 2026, with MIT-licensed weights, trai…
The U.S. Commerce Department sent an enforcement letter invoking an export-control directive that barred non-U.S. access to Anthropic's latest AI models, Fable 5 and Mythos 5, leading Anthropic to disable access to compl…
PeekAI, a new open-source tool for local-first observability and debugging of Python AI agents, launched on Hacker News. The tool stores traces in SQLite, requires no cloud accounts or API keys, and supports multi-agent …
AI agents struggle to develop expertise in new domains from text corpora alone, relying on shallow methods like RAG or long-context processing. Researchers propose 'Machine Studying' as the problem of autonomously turnin…
Lithp.py, a Lisp interpreter for John McCarthy's original Lisp, was released around 2008 by developer Fogus. The interpreter, written in Python 2.6.1+, includes core functions like car, cdr, cons, and lambda, and provide…
Large language models are stateless by default, but agents require memory to retain context across steps. A new technical guide identifies seven types of agent memory—working, semantic, episodic, procedural, retrieval, p…
Norway announced a near-total ban on generative AI for children ages 6 to 13 in schools, effective August 2026, citing risks to cognitive development and data privacy. Prime Minister Jonas Gahr Støre stated the policy ai…
A developer argues that AI agents are defined by their harness—the software infrastructure wrapping a model—not the model itself. The thesis states that as harness engineering becomes a distinct discipline, it will absor…
Samsung Electronics is rolling out ChatGPT Enterprise and Codex to employees worldwide, reversing its 2023 ban on generative AI after a data leak incident. The South Korean tech giant becomes one of OpenAI's largest ente…
ByteDance released DeerFlow 2.0, a ground-up rewrite of its open-source agent runtime, on February 28, 2026. The new version transforms the deep-research tool into an isolated, stateful execution harness for autonomous s…
Google announced on May 19, 2026, that it is retiring its open-source Gemini CLI and replacing it with Antigravity CLI, a Go-based, agent-first platform. The migration, effective June 18, 2026, introduces breaking change…
The Trump administration imposed a de facto ban on Anthropic's Mythos derivative Fable 5, citing national security concerns and export controls, forcing Anthropic to disband both models for all customers. The move has sp…
A developer fine-tuned a tiny 0.6B-parameter Qwen 3 model to categorize household questions into metadata categories like pool, car, and HVAC. The baseline model achieved only 10% accuracy via prompting, but fine-tuning …
Andrew Curran claimed Sunday that a more capable version of Anthropic's Mythos-class model has emerged from training, nine days after the US government forced Anthropic to disable Fable 5 and Mythos 5. The claim, unconfi…
The Swiss AI Initiative released Apertus Mini on June 15, a set of 16 smaller language models distilled from the open-source Apertus foundation model, designed for local deployment on constrained hardware. The move aims …
A new analysis argues that calling large language models like ChatGPT 'autocomplete' is technically accurate but deeply misleading, because the models build a high-dimensional internal state via attention mechanisms befo…