All tags
Company: "goodfireai"
not much happened today
raev2 gated-deltanet-2 kda mamba-3 dclm nvidia openai nous-research representation-learning tokenization linear-attention long-context mechanistic-interpretability math data-filtering agent-infrastructure language-modeling commonsense-reasoning 1jaskiratsingh recatm sainingxie ahatamiz1 rasbt nousresearch tatsu_hashimoto goodfireai markchen90 wtgowers memecrashes cloneofsimo lvwerra
RAEv2 advances representation-first tokenization with >10x faster convergence and improved generation, tested on text-to-image and world models. NVIDIA's Gated DeltaNet-2 innovates linear attention with channel-wise gates, outperforming KDA and Mamba-3 at 1.3B parameters on language modeling and reasoning tasks. Studies on subword tokenization reveal only some benefits at scale, while data filtering research suggests that with enough compute, no filtering may be optimal at around 1e30 FLOPs. Mechanistic interpretability updates propose clustering features by joint firing patterns for better geometry understanding. OpenAI's AI-assisted breakthrough on an ErdΕs unit-distance math problem sparks debate on AI's role in mathematical research. Harnesses remain key for capability improvements in agent infrastructure.
GPT-Realtime-2, -Translate, and -Whisper: new SOTA realtime voice APIs
gpt-realtime-2 gpt-5.5 codex openai anthropic goodfireai scale-ai voice-models streaming-translation transcription benchmarking context-windows browser-automation cybersecurity interpretability neural-geometry manifolds ai-safety rlhf micahcarroll milesbrundage ryanpgreenblatt
OpenAI released GPT-Realtime-2, a voice model with GPT-5-class reasoning, tool use, interruption handling, and extended context windows up to 128K tokens, achieving top scores on Big Bench Audio and Conversational Dynamics benchmarks. They also launched a Chrome plugin for Codex enabling browser control and multitasking, and introduced GPT-5.5 with Trusted Access for Cyber for secure defensive workflows and red teaming. Anthropic introduced Natural Language Autoencoders for interpreting model activations as human-readable text, aiding interpretability and debugging, while Goodfire proposed a neural geometry research agenda focusing on manifolds as primitives for neural network behavior. Anthropic also announced The Anthropic Institute to advance AI safety and economic resilience research.
Cursor 2.0 & Composer-1: Fast Models and New Agents UI
composer-1 gpt-oss-safeguard-20b gpt-oss-safeguard-120b gpt-oss gpt-5-mini cursor_ai openai huggingface ollama cerebras groq goodfireai rakuten agentic-coding reinforcement-learning mixture-of-experts fine-tuning policy-classification open-weight-models inference-stacks cost-efficiency multi-agent-systems ide voice-to-code code-review built-in-browser model-optimization sasha_rush dan_shipper samkottler ellev3n11 swyx
Cursor 2.0 launched with Composer-1, an agentic coding model optimized for speed and precision, featuring multi-agent orchestration, built-in browser for testing, and voice-to-code capabilities. OpenAI released gpt-oss-safeguard models (20B, 120B) for policy-based safety classification, open-weight and fine-tuned from gpt-oss, available on Hugging Face and supported by inference stacks like Ollama and Cerebras. Goodfire and Rakuten demonstrated sparse autoencoders for PII detection matching gpt-5-mini accuracy at significantly lower cost. The Cursor 2.0 update also includes a redesigned interface for managing multiple AI coding agents, marking a major advancement in AI IDEs. "Fast-not-slowest" tradeoff emphasized by early users for Composer-1, enabling rapid iteration with human-in-the-loop.
Grok 3 & 3-mini now API Available
grok-3 grok-3-mini gemini-2.5-flash o3 o4-mini llama-4-maverick gemma-3-27b openai llamaindex google-deepmind epochairesearch goodfireai mechanize agent-development agent-communication cli-tools reinforcement-learning model-evaluation quantization-aware-training model-compression training-compute hybrid-reasoning model-benchmarking
Grok 3 API is now available, including a smaller version called Grok 3 mini, which offers competitive pricing and full reasoning traces. OpenAI released a practical guide for building AI agents, while LlamaIndex supports the Agent2Agent protocol for multi-agent communication. Codex CLI is gaining traction with new features and competition from Aider and Claude Code. GoogleDeepMind launched Gemini 2.5 Flash, a hybrid reasoning model topping the Chatbot Arena leaderboard. OpenAI's o3 and o4-mini models show emergent behaviors from large-scale reinforcement learning. EpochAIResearch updated its methodology, removing Maverick from high FLOP models as Llama 4 Maverick training compute drops. GoodfireAI announced a $50M Series A for its Ember neural programming platform. Mechanize was founded to build virtual work environments and automation benchmarks. GoogleDeepMind's Quantisation Aware Training for Gemma 3 models reduces model size significantly, with open source checkpoints available.