All tags
Person: "gdb"
OpenAI Titan XPU: 10GW of self-designed chips with Broadcom
llama-3-70b openai nvidia amd broadcom inferencemax asic inference compute-infrastructure chip-design fp8 reinforcement-learning ambient-agents custom-accelerators energy-consumption podcast gdb
OpenAI is finalizing a custom ASIC chip design to deploy 10GW of inference compute, complementing existing deals with NVIDIA (10GW) and AMD (6GW). This marks a significant scale-up from OpenAI's current 2GW compute, aiming for a roadmap of 250GW total, which is half the energy consumption of the US. Greg from OpenAI highlights the shift of ChatGPT from interactive use to always-on ambient agents requiring massive compute, emphasizing the challenge of building chips for billions of users. The in-house ASIC effort was driven by the need for tailored designs after limited success influencing external chip startups. Broadcom's stock surged 10% on the news. Additionally, InferenceMAX reports improved ROCm stability and nuanced performance comparisons between AMD MI300X and NVIDIA H100/H200 on llama-3-70b FP8 workloads, with RL training infrastructure updates noted.
OpenAI Dev Day: Apps SDK, AgentKit, Codex GA, GPT‑5 Pro and Sora 2 APIs
gpt-5-pro gpt-realtime-mini-2025-10-06 gpt-audio-mini-2025-10-06 gpt-image-1-mini sora-2 sora-2-pro openai canva figma zillow coursera api model-release fine-tuning agentic-ai code-generation model-deployment pricing prompt-optimization software-development multimodality sama edwinarbus gdb dbreunig stevenheidel
OpenAI showcased major product launches at their DevDay including the Apps SDK, AgentKit, and Codex now generally available with SDK and enterprise features. They introduced new models such as gpt-5-pro, gpt-realtime-mini-2025-10-06, gpt-audio-mini-2025-10-06, gpt-image-1-mini, and sora-2 with a pro variant. The Apps SDK enables embedding interactive apps inside ChatGPT with partners like Canva, Figma, Zillow, and Coursera. AgentKit offers a full stack for building and deploying production agents with tools like ChatKit and Guardrails. Codex supports speech and controller-driven coding, credited with high internal shipping velocity. Pricing for GPT-5 Pro was revealed at $15 input and $120 output per million tokens. "OpenAI turned ChatGPT into an application platform" and "AgentKit built a working agent in under 8 minutes" were highlights.
GDPVal finding: Claude Opus 4.1 within 95% of AGI (human experts in top 44 white collar jobs)
claude-4.1-opus gpt-5-high gptnext gemini-2.5-flash gemini-2.5-flash-lite deepseek-v3.1-terminus google-chirp-2 qwen-2.5b openai anthropic google nvidia artificial-analysis deepseek benchmarking agentic-ai tool-use long-context speech-to-text model-evaluation reasoning pricing model-performance kevinweil gdb dejavucoder yuchenj_uw lhsummers
OpenAI's Evals team released GDPval, a comprehensive evaluation benchmark covering 1,320 tasks across 44 predominantly digital occupations, assessing AI models against human experts with 14 years average experience. Early results show Claude 4.1 Opus outperforming human experts in most categories and GPT-5 high trailing behind, with projections that GPTnext could match human performance by mid-2026. The benchmark is positioned as a key metric for policymakers and labor impact forecasting. Additionally, Artificial Analysis reported improvements in Gemini 2.5 Flash/Flash-Lite and DeepSeek V3.1 Terminus models, alongside new speech-to-text benchmarks (AA-WER) highlighting leaders like Google Chirp 2 and NVIDIA Canary Qwen2.5B. Agentic AI advances include Kimi OK Computer, an OS-like agent with extended tool capabilities and new vendor verification tools.
not much happened today
qwen3-max qwen3-vl qwen3-coder-plus gpt-5-codex code-world-model-32b claude-sonnet-4 claude-opus-4.1 alibaba openai meta-ai-fair huggingface anthropic microsoft github context-windows code-generation model-releases model-benchmarking api model-optimization multimodality software-engineering model-training huybery akhaliq lmarena_ai gdb ylecun pierceboggan julesagent
Alibaba unveiled the Qwen3 model family including Qwen3-Max and Qwen3-VL with a native 256K context window expandable to 1M, strong OCR in 32 languages, and rapid release velocity (~3.5 releases/month) backed by a $52B infrastructure roadmap. OpenAI launched GPT-5 Codex, an agent-optimized coding model with up to 400K context and adaptive reasoning priced at $1.25/$10 per million tokens, integrated into Cline and benchmarked in WebDev arenas. Meta AI FAIR released the open-weight Code World Model (CWM) 32B, a dense code generation model with strong benchmark scores (e.g., 65.8% SWE-bench Verified, 96.6% Math-500) and public safety reports. Ecosystem updates include GitHub Copilot's new embedding model for faster code search and Anthropic's Claude Sonnet 4 and Opus 4.1 integration into Microsoft 365 Copilot. The vLLM 0.10.2 update introduces Decode Context Parallel (DCP) for improved system performance.
NVIDIA to invest $100B in OpenAI for 10GW of Vera Rubin rollout
qwen3-omni deepseek-v3.1 nvidia openai oracle intel enfabrica wayne gpu-infrastructure deterministic-inference reinforcement-learning fp8-precision gpu-performance ai-infrastructure strategic-partnerships investment datacenters cuda-graphs pipeline-parallelism data-parallelism artificialanlys gdb
NVIDIA and OpenAI announced a landmark strategic partnership to deploy at least 10 gigawatts of AI datacenters using NVIDIA's systems, with NVIDIA investing up to $100 billion progressively as each gigawatt is deployed, starting in the second half of 2026 on the Vera Rubin platform. This deal significantly impacts the AI infrastructure funding landscape, potentially supporting OpenAI's $300 billion commitment to Oracle. The announcement caused major stock market reactions, with NVIDIA's market cap surging by $170 billion. Additionally, advancements in deterministic inference for reinforcement learning and FP8 precision gains in GPU performance were highlighted by AI practitioners.
not much happened today
gpt-5-codex vllm-0.10.2 qwen3-next-80b hunyuanimage-2.1 openai microsoft perplexity-ai huggingface amd tencent lmstudio agentic-ai ide context-windows inference distributed-inference reinforcement-learning robotics long-context model-optimization text-to-image multimodality model-licenses gdb teknium1 finbarrtimbers thsottiaux theturingpost pierceboggan amandaksilver aravsrinivas sergiopaniego art_zucker danielhanchen rwojo awnihannun
GPT-5 Codex rollout shows strong agentic coding capabilities with some token bloat issues. IDEs like VS Code Insiders and Cursor 1.6 enhance context windows and model integration. vLLM 0.10.2 supports aarch64 and NVIDIA GB200 with performance improvements. AMD ROCm updates add modern attention, sparse MoE, and distributed inference. TRL introduces Context Parallelism for long-context training. Robotics and RL data pipelines improve with Unsloth and LeRobotDataset v3. Qwen3-Next-80B runs efficiently on Mac M4 Max with MLX. Tencent's HunyuanImage 2.1 is a 17B bilingual text-to-image model with 2048×2048 resolution and restricted open weights.
not much happened today
gpt-5 grok-code-fast-1 claude-sonnet glm-4.5 longcat-flash-chat fastvlm mobileclip2 internvl3.5 openai x-ai zhipu-ai meituan apple model-architecture moe adaptive-compute inference-speed model-training cost-efficiency coding developer-tools open-inference on-device-ai vision gdb martin_casado yanndubs elonmusk cline vikhyatk dzhng quixiai tim_dettmers casper_hansen_ reach_vb eliebakouch teortaxestex youjiacheng
OpenAI integrates GPT-5 into Xcode 26 with improved coding latency, though some UX trade-offs are noted. xAI's Grok Code Fast 1 gains momentum, surpassing Claude Sonnet in usage and praised for fast debugging. Zhipu's GLM-4.5 offers a cost-effective coding plan with strong performance against Claude Sonnet 4. Meituan releases the LongCat-Flash-Chat, a 560B parameter MoE model with adaptive compute and detailed technical insights. Apple debuts on-device vision-language models FastVLM and MobileCLIP2 alongside InternVL3.5.
not much happened today
fastvlm mobileclip2 grok-code-fast-1 gpt-5 qwen-3-coder-30b-a3b apple hugging-face x-ai openai groq run-llama lmstudio vision model-quantization code-generation cli-workflows retrieval-augmentation embedding-models local-ai multimodality reach_vb xenovacom pcuenq awnihannun cline veggie_eric nickbaumann_ gdb benankdev loganmarkewich tom_doerr fastmcp ggerganov orionweller antoine_chaffin
Apple released three real-time vision-language models (FastVLM, MobileCLIP2) on Hugging Face with significant speed and size improvements, supporting WebGPU and Core ML. Their MLX framework now supports MXFP4 format, competing with NVFP4 for FP4 quantization. xAI launched grok-code-fast-1, outperforming Claude for code edits, while OpenAI integrated GPT-5 into Xcode 26 and released a new Responses API on Groq hardware. CLI-first agent workflows advanced with tools like SemTools, MLX local runner for Apple Silicon, and llama.vim recommending Qwen 3 Coder 30B A3B. Retrieval research highlights limitations of single-vector embeddings, promoting ColBERT-style late interaction.
OpenAI Realtime API GA and new `gpt-realtime` model, 20% cheaper than 4o
gpt-realtime gpt-4o-realtime grok-code-fast-1 codex mai-1-preview mai-voice-1 gemini-cli openai xai microsoft google speech-to-speech instruction-following function-calling telephony webrtc voice-agents multilingual-switching voice-control benchmarks coding-models ide-integration developer-tools model-updates swyx juberti omarsar0 reach_vb pbbakkum skcd42 mohitreddy13 cline kevinweil gdb sama _philschmid
OpenAI launched the gpt-realtime model and Realtime API to GA, featuring advanced speech-to-speech capabilities, new voices (Cedar, Marin), image input, SIP telephony, and a ~20% price cut. Benchmarks show improvements over gpt-4o-realtime on BigBench and ComplexFuncBench. xAI introduced Grok Code Fast 1, a speed-optimized coding model integrated with popular IDEs, while OpenAI Codex received major upgrades for local and cloud development workflows. Google’s Gemini CLI improved multi-editor support, and new models like Microsoft MAI-1-preview and MAI-Voice-1 were announced. "The new all-in-one WebRTC API removes the ephemeral token step and supports video on the same connection," highlighting enhanced developer tooling.
not much happened today
grok-2 grok-2.5 vibevoice-1.5b motif-2.6b gpt-5 qwen-code xai-org microsoft motif-technology alibaba huggingface langchain-ai mixture-of-experts model-scaling model-architecture text-to-speech fine-tuning training-data optimization reinforcement-learning agentic-ai tool-use model-training model-release api software-development model-quantization elonmusk clementdelangue rasbt quanquangu akhaliq eliebakouch gdb ericmitchellai ivanfioravanti deanwball giffmana omarsar0 corbtt
xAI released open weights for Grok-2 and Grok-2.5 with a novel MoE residual architecture and μP scaling, sparking community excitement and licensing concerns. Microsoft open-sourced VibeVoice-1.5B, a multi-speaker long-form TTS model with streaming support and a 7B variant forthcoming. Motif Technology published a detailed report on Motif-2.6B, highlighting Differential Attention, PolyNorm, and extensive finetuning, trained on AMD MI250 GPUs. In coding tools, momentum builds around GPT-5-backed workflows, with developers favoring it over Claude Code. Alibaba released Qwen-Code v0.0.8 with deep VS Code integration and MCP CLI enhancements. The MCP ecosystem advances with LiveMCP-101 stress tests, the universal MCP server "Rube," and LangGraph Platform's rollout of revision queueing and ART integration for RL training of agents.
Databricks' $100B Series K
deepseek-v3.1-base deepseek-v3.1-instruct chatgpt-go qwen-image-edit databricks openai deepseek hugging-face alibaba model-release benchmarking pricing-models fine-tuning model-architecture image-editing video-generation api agentic-ai sama nickaturley kevinweil gdb sherwinwu nptacek reach_vb clementdelangue teortaxestex quixiai georgejrjrjr scaling01 alibaba_qwen linoy_tsaban ostrisai lmarena_ai
Databricks reached a $100 billion valuation, becoming a centicorn with new Data (Lakebase) and AI (Agent Bricks) products. OpenAI launched ChatGPT Go in India at ₹399/month (~$4.55), offering significantly increased usage limits and UPI payment support, with plans for global expansion. The DeepSeek V3.1 Base/Instruct models were quietly released on Hugging Face, showing strong coding benchmark performance and adopting an Anthropic-style hybrid system. The Qwen-Image-Edit model from Alibaba is gaining traction with integrations and community pruning experiments. "DeepSeek V3.1 Base outperforms Claude 4 Opus on coding benchmarks" and "ChatGPT Go offers 10x higher message limits and 2x longer memory" highlight key advancements.
not much happened today
gpt-5 gpt-5-high gpt-5-mini-high gpt-5-nano-high imagen-4 gemma-3-270m openai google lmsys model-releases model-performance prompt-engineering developer-tools image-generation model-optimization transformers tokenization model-scaling sama aidan_mclau kevinweil lmarena_ai edwinarbus gdb omarsar0 philschmid m4rkmc
OpenAI rolled out GPT-5 as the default in ChatGPT with new modes and a "warmer" personality, plus expanded message limits for Plus/Team users and Enterprise/Edu access. Performance rankings show gpt-5-high leading, with smaller variants also ranked, though critiques note some underperformance versus Chinese models and sensitivity to sycophancy. OpenAI enhanced developer tools with a "Quick eval" feature, coding tips, and an improved Playground. Google released Imagen 4 generally available with faster generation and higher resolution, plus the ultra-small Gemma 3 270M model with a large vocabulary and ecosystem support. Podcasts featured OpenAI leaders discussing GPT-5 systems, routing, and efficiency.
OpenAI's IMO Gold model also wins IOI Gold
gpt-5 gpt-5-thinking gpt-5-mini gemini-2.5-pro claude opus-4.1 openai google-deepmind anthropic reinforcement-learning benchmarking model-performance prompt-engineering model-behavior competitive-programming user-experience model-naming model-selection hallucination-detection sama scaling01 yanndubs sherylhsu ahmed_el-kishky jerry_tworek noam_brown alex_wei amandaaskell ericmitchellai jon_durbin gdb jerryjliu0
OpenAI announced placing #6 among human coders at the IOI, reflecting rapid progress in competitive coding AI over the past two years. The GPT-5 launch faced significant user backlash over restrictive usage limits and removal of model selection control, leading to a reversal and increased limits to 3000 requests per week for Plus users. Confusion around GPT-5 naming and benchmarking was highlighted, with critiques on methodological issues comparing models like Claude and Gemini. Performance reviews of GPT-5 are mixed, with claims of near-zero hallucinations by OpenAI staff but user reports of confidence in hallucinations and steering difficulties. Benchmarks show GPT-5 mini performing well on document understanding, while the full GPT-5 is seen as expensive and middling. On the Chatbot Arena, Gemini 2.5 Pro holds a 67% winrate against GPT-5 Thinking. Prompting and model behavior remain key discussion points.
not much happened today
gpt-5 gpt-4o grok-4 claude-4-sonnet openai microsoft reasoning latency model-routing benchmarking reinforcement-learning hallucination-control creative-writing priority-processing api-traffic model-deprecation user-experience model-selection voice-mode documentation sama nickaturley elaineyale6 scaling01 mustafasuleyman kevinweil omarsar0 jeremyphoward juberti epochairesearch lechmazur gdb
OpenAI launched GPT-5 with a unified user experience removing manual model selection, causing initial routing and access issues for Plus users that are being addressed with fixes including restored model options and increased usage limits. GPT-5 introduces "Priority Processing" for lower latency at higher price tiers, achieving ~750ms median time-to-first-token in some cases. Microsoft reports full Copilot adoption of GPT-5, and API traffic doubled within 24 hours, peaking at 2 billion tokens per minute. Early benchmarks show GPT-5 leading in reasoning tasks like FrontierMath and LiveBench, with improvements in hallucination control and creative writing, though some models like Grok-4 and Claude-4 Sonnet Thinking outperform it in specific RL-heavy reasoning benchmarks. OpenAI also released extensive migration and feature guides but faced some rollout issues including a broken code sample and a problematic Voice Mode launch. "Unified GPT-5" ends model pickers, pushing developers away from manual model selection.
not much happened today
glm-4.5 glm-4.5-air qwen3-coder qwen3-235b kimi-k2 grok-imagine wan-2.2 smollm3 figure-01 figure-02 vitpose++ chatgpt zhipu-ai alibaba moonshot-ai x-ai figure openai runway mlx ollama deeplearningai model-releases model-performance moe image-generation video-generation pose-estimation robotics training-code-release interactive-learning in-context-learning yuchenj_uw corbtt reach_vb ollama deeplearningai gdb sama c_valenzuelab adcock_brett skalskip92 loubnabenallal1 hojonathanho ostrisai
Chinese AI labs have released powerful open-source models like GLM-4.5 and GLM-4.5-Air from Zhipu AI, Qwen3 Coder and Qwen3-235B from Alibaba, and Kimi K2 from Moonshot AI, highlighting a surge in permissively licensed models. Zhipu AI's GLM-4.5 is a 355B parameter MoE model competitive with Claude 4 Opus and Gemini 2.5 Pro. Alibaba's Qwen3 Coder shows strong code generation performance with a low edit failure rate, while Moonshot AI's Kimi K2 is a 1 trillion-parameter MoE model surpassing benchmarks like LiveCodeBench. In video and image generation, xAI launched Grok Imagine, and Wan2.2 impressed with innovative image-to-video generation. Robotics advances include Figure's Figure-01 and Figure-02 humanoid robots and ViTPose++ for pose estimation in basketball analysis. SmolLM3 training and evaluation code was fully released under Apache 2.0. OpenAI introduced Study Mode in ChatGPT to enhance interactive learning, and Runway rolled out Runway Aleph, a new in-context video model for multi-task visual generation. The community notes a competitive disadvantage for organizations avoiding these Chinese open-source models. "Orgs avoiding these models are at a significant competitive disadvantage," noted by @corbtt.
ChatGPT Agent: new o* model + unified Deep Research browser + Operator computer use + Code Interpreter terminal
o3 o4 gptnext openai reinforcement-learning benchmarking model-performance model-risk long-context model-deployment fine-tuning sama gdb kevinweil xikun_zhang_ keren_gu boazbaraktcs
OpenAI launched the ChatGPT Agent, a new advanced AI system capable of browsing the web, coding, analyzing data, and creating reports, marking a significant step towards human-like computer use. The agent, distinct from and superior to o3, is considered the first public exposure of what was internally called o4, now merged into GPTNext. It features end-to-end reinforcement learning, can operate for extended periods (tested up to 2 hours), and is classified as "High" risk for biological misuse, with safeguards activated. Early benchmarks show mixed results, excelling in some tests like WebArena and BrowserComp but underperforming on others like PaperBench. Key figures involved include Sam Altman, Greg Brockman, and Kevin Weil, with technical insights from xikun_zhang_ and risk commentary from KerenGu and boazbaraktcs. The launch sparked speculation about GPT-5, which was confirmed not to be the case.
Execuhires Round 2: Scale-Meta, Lamini-AMD, and Instacart-OpenAI
o3-pro o3 o1-pro gpt-4o gpt-4.1 gpt-4.1-mini gpt-4.1-nano meta-ai-fair scale-ai lamini amd openai gemini google anthropic model-release benchmarking reasoning fine-tuning pricing model-performance direct-preference-optimization complex-problem-solving alexandr_wang sharon_zhou fidji_simo sama jack_rae markchen90 kevinweil gdb gregkamradt lechmazur wesrothmoney paul_cal imjaredz cto_junior johnowhitaker polynoamial scaling01
Meta hires Scale AI's Alexandr Wang to lead its new "Superintelligence" division following a $15 billion investment for a 49% stake in Scale. Lamini's Sharon Zhou joins AMD as VP of AI under Lisa Su, while Instacart's Fidji Simo becomes CEO of Apps at OpenAI under Sama. Meta offers over $10 million/year compensation packages to top researchers, successfully recruiting Jack Rae from Gemini. OpenAI releases o3-pro model to ChatGPT Pro users and API, outperforming o3 and setting new benchmarks like Extended NYT Connections and SnakeBench. Despite being slower than o1-pro, o3-pro excels in reasoning and complex problem-solving. OpenAI cuts o3 pricing by 80%, making it cheaper than GPT-4o and pressuring competitors like Google and Anthropic to lower prices. Users can now fine-tune the GPT-4.1 family using direct preference optimization (DPO) for subjective tasks.
Reasoning Price War 2: Mistral Magistral + o3's 80% price cut + o3-pro
o3 o3-pro gpt-4.1 claude-4-sonnet gemini-2.5-pro magistral-small magistral-medium mistral-small-3.1 openai anthropic google-deepmind mistral-ai perplexity-ai reasoning token-efficiency price-cut benchmarking open-source model-releases context-windows gpu-optimization swyx sama scaling01 polynoamial nrehiew_ kevinweil gdb flavioad stevenheidel aravsrinivas
OpenAI announced an 80% price cut for its o3 model, making it competitively priced with GPT-4.1 and rivaling Anthropic's Claude 4 Sonnet and Google's Gemini 2.5 Pro. Alongside, o3-pro was released as a more powerful and reliable variant, though early benchmarks showed mixed performance relative to cost. Mistral AI launched its Magistral reasoning models, including an open-source 24B parameter version optimized for efficient deployment on consumer GPUs. The price reduction and new model releases signal intensified competition in reasoning-focused large language models, with notable improvements in token efficiency and cost-effectiveness.
Apple exposes Foundation Models API and... no new Siri
chatgpt apple openai langchain llamaindex on-device-ai foundation-models reasoning reinforcement-learning voice translation software-automation agentic-workflows gdb scaling01 giffmana kevinweil
Apple released on-device foundation models for iOS developers, though their recent "Illusion of Reasoning" paper faced significant backlash for flawed methodology regarding LLM reasoning. OpenAI updated ChatGPT's Advanced Voice Mode with more natural voice and improved translation, demonstrated by Greg Brockman. LangChain and LlamaIndex launched new AI agents and tools, including a SWE Agent for software automation and an Excel agent using reinforcement learning for data transformation. The AI community engaged in heated debate over reasoning capabilities of LLMs, highlighting challenges in evaluation methods.
not much happened today
codex claude-4-opus claude-4-sonnet gemini-2.5-pro gemini-2.5 qwen-2.5-vl qwen-3 playdiffusion openai anthropic google perplexity-ai bing playai suno hugging-face langchain-ai qwen mlx assemblyai llamacloud fine-tuning model-benchmarking text-to-video agentic-ai retrieval-augmented-generation open-source-models speech-editing audio-processing text-to-speech ultra-low-latency multimodality public-notebooks sama gdb kevinweil lmarena_ai epochairesearch reach_vb wightmanr deeplearningai mervenoyann awnihannun jordirib1 aravsrinivas omarsar0 lioronai jerryjliu0 nerdai tonywu_71 _akhaliq clementdelangue _mfelfel
OpenAI rolled out Codex to ChatGPT Plus users with internet access and fine-grained controls, improving memory features for free users. Anthropic's Claude 4 Opus and Sonnet models lead coding benchmarks, while Google's Gemini 2.5 Pro and Flash models gain recognition with new audio capabilities. Qwen 2.5-VL and Qwen 3 quantizations are noted for versatility and support. Bing Video Creator launched globally enabling text-to-video generation, and Perplexity Labs sees increased demand for travel search. New agentic AI tools and RAG innovations include LlamaCloud and FedRAG. Open-source releases include Holo-1 for web navigation and PlayAI's PlayDiffusion for speech editing. Audio and multimodal advances feature Suno's music editing upgrades, Google's native TTS in 24+ languages, and Universal Streaming's ultra-low latency speech-to-text. Google NotebookLM now supports public notebooks. "Codex's internet access brings tradeoffs, with explicit warnings about risk" and "Gemini 2.5 Pro is cited as a daily driver by users".
Gemini's AlphaEvolve agent uses Gemini 2.0 to find new Math and cuts Gemini cost 1% — without RL
gemini gpt-4.1 gpt-4o-mini o3 o4-mini google-deepmind openai algorithm-discovery coding-agents matrix-multiplication optimization reinforcement-learning model-weights training-efficiency safety-evaluations instruction-following coding-tasks model-releases _philschmid scott_swingle alex_dimakis henry jason_wei kevinweil michpokrass scaling01 gdb
Deepmind's AlphaEvolve, a 2025 update to AlphaTensor and FunSearch, is a Gemini-powered coding agent for algorithm discovery that designs faster matrix multiplication algorithms, solves open math problems, and improves data center and AI training efficiency. It achieves a 23% faster kernel speedup in Gemini training and surpasses state-of-the-art on 20% of applied problems, including improvements on the Minimum Overlap Problem and Kissing number problem. Unlike Deep-RL, it optimizes code pieces rather than model weights. Meanwhile, OpenAI released GPT-4.1 in ChatGPT, specializing in coding and instruction following, with a faster alternative GPT-4.1 mini replacing GPT-4o mini for all users. OpenAI also launched the Safety Evaluations Hub and the OpenAI to Z Challenge using o3/o4 mini and GPT-4.1 models to discover archaeological sites. "Maybe midtrain + good search is all you need for AI for scientific innovation" - Jason Wei.
not much happened today
hunyuan-turbos qwen3-235b-a22b o3 gpt-4.1-nano grok-3 gemini-2.5-pro seed1.5-vl kling-2.0 tencent openai bytedance meta-ai-fair nvidia deepseek benchmarking model-performance moe reasoning vision video-understanding vision-language multimodality model-evaluation model-optimization lmarena_ai artificialanlys gdb _jasonwei iScienceLuvr _akhaliq _philschmid teortaxesTex mervenoyann reach_vb
Tencent's Hunyuan-Turbos has risen to #8 on the LMArena leaderboard, showing strong performance across major categories and significant improvement since February. The Qwen3 model family, especially the Qwen3 235B-A22B (Reasoning) model, is noted for its intelligence and efficient parameter usage. OpenAI introduced HealthBench, a new health evaluation benchmark developed with input from over 250 physicians, where models like o3, GPT-4.1 nano, and Grok 3 showed strong results. ByteDance released Seed1.5-VL, a vision-language model with a 532M-parameter vision encoder and a 20B active parameter MoE LLM, achieving state-of-the-art results on 38 public benchmarks. In vision-language, Kling 2.0 leads image-to-video generation, and Gemini 2.5 Pro excels in video understanding with advanced multimodal capabilities. Meta's Vision-Language-Action framework and updates on VLMs for 2025 were also highlighted.
OpenAI o3, o4-mini, and Codex CLI
o3 o4-mini gemini-2.5-pro claude-3-sonnet chatgpt openai reinforcement-learning performance vision tool-use open-source coding-agents model-benchmarking multimodality scaling inference sama aidan_mclau markchen90 gdb aidan_clark_ kevinweil swyx polynoamial scaling01
OpenAI launched the o3 and o4-mini models, emphasizing improvements in reinforcement-learning scaling and overall efficiency, making o4-mini cheaper and better across prioritized metrics. These models showcase enhanced vision and tool use capabilities, though API access for these features is pending. The release includes Codex CLI, an open-source coding agent that integrates with these models to convert natural language into working code. Accessibility extends to ChatGPT Plus, Pro, and Team users, with o3 being notably more expensive than Gemini 2.5 Pro. Performance benchmarks highlight the intelligence gains from scaling inference, with comparisons against models like Sonnet and Gemini. The launch has been well received despite some less favorable evaluation results.
Learnings from o1 AMA
o1-preview o1-mini claude-3.5-sonnet gpt-4o openai weights-biases cohere weaviate reinforcement-learning chain-of-thought reasoning model-performance prompting code-editing rag hybrid-search sama rohanpaul_ai gdb andrew-mayne
OpenAI released the o1 model series, touted as their "most capable and aligned models yet," trained with reinforcement learning to enhance reasoning. The o1-preview model scored 21% on ARC-AGI, ~80% on aider code editing (surpassing Claude 3.5 Sonnet's 77%), and ~52% on Cognition-Golden, showcasing a shift from memorizing answers to memorizing reasoning. The model employs a unique chain-of-thought approach enabling "System II thinking" for better problem-solving. Experts like Andrew Mayne advise framing o1 as a smart friend providing thoughtful explanations. Additionally, an advanced RAG course sponsored by Weights & Biases, Cohere, and Weaviate offers strategies for hybrid search and prompting to optimize AI solutions.
Francois Chollet launches $1m ARC Prize
gpt-4 chatgpt openai apple togethercompute benchmarking agi pattern-recognition skill-acquisition privacy on-device-ai mixed-precision-quantization mixture-of-experts multimodality agentic-ai francois-chollet karpathy svpino philschmid clementdelangue sama gdb miramurati kevin-weil sarah-friar
François Chollet critiques current paths to AGI, emphasizing the importance of benchmarks that resist saturation and focus on skill acquisition and open-ended problem solving. The ARC-AGI puzzles exemplify "easy for humans, hard for AI" challenges to measure progress toward AGI. Meanwhile, Apple announces integration of ChatGPT into iOS, iPadOS, and macOS through a partnership with OpenAI, enabling AI-powered features like document summarization and photo analysis with privacy-preserving measures. Discussions highlight Apple's focus on deep AI integration and on-device models optimized with techniques like mixed-precision quantization, though some skepticism remains about their AI capabilities compared to GPT-4. Additionally, Together Compute introduces a Mixture of Agents approach achieving strong performance on AlpacaEval 2.0.
Qwen 2 beats Llama 3 (and we don't know how)
qwen-2 llama-3 llama-3-70b gpt-4 nllb alibaba groq meta-ai-fair multilinguality benchmarking inference-speed sparse-autoencoders scaling-laws post-training instruction-following rejection-sampling execution-feedback model-release multilingual-models model-training philschmid huybery jonathanross321 awnihannun gdb nabla_theta ylecun
Alibaba released Qwen 2 models under Apache 2.0 license, claiming to outperform Llama 3 in open models with multilingual support in 29 languages and strong benchmark scores like MMLU 82.3 and HumanEval 86.0. Groq demonstrated ultra-fast inference speed on Llama-3 70B at 40,792 tokens/s and running 4 Wikipedia articles in 200ms. Research on sparse autoencoders (SAEs) for interpreting GPT-4 neural activity showed new training methods, metrics, and scaling laws. Meta AI announced the No Language Left Behind (NLLB) model capable of high-quality translations between 200 languages, including low-resource ones. "Our post-training phase is designed with the principle of scalable training with minimal human annotation," highlighting techniques like rejection sampling for math and execution feedback for coding.
GPT-4o: the new SOTA-EVERYTHING Frontier model (GPT4O version)
gpt-4o gpt-4-turbo openai lmsys multion adept multimodality vision speech-recognition tokenization real-time-processing coding model-performance model-optimization desktop-agents sama gdb
OpenAI has released GPT-4o, a new multimodal model capable of reasoning across text, audio, and video in real time with low latency (~300ms). It features voice and vision capabilities, improved non-English language performance with an expanded 200k vocabulary tokenizer, and is available to all ChatGPT users including free plans. GPT-4o is half the price and twice as fast as GPT-4-turbo with 5x rate limits. The model supports real-time voice and video input/output and shows strong coding capabilities. The release includes a new desktop app that can read screen and clipboard history, challenging existing desktop agent startups. The announcement was accompanied by demos including image generation and 3D object handling, with OpenAI achieving state-of-the-art performance in ASR and vision tasks. The update was widely discussed on social media, with comparisons to GPT-4T highlighting GPT-4o's speed and versatility. "GPT-4o is smart, fast, natively multimodal, and a step towards more natural human-computer interaction" and "extremely versatile and fun to play with".
Quis promptum ipso promptiet?
llama-3-70b llama-3-120b llama-3 llama-cpp anthropic openai zoominfo neuralink prompt-engineering chain-of-thought rag quantization cuda-graphs gpu-optimization thought-controlled-devices modeling-consciousness conference sama gdb bindureddy svpino rohanpaul_ai alexalbert__ abacaj
Anthropic released upgrades to their Workbench Console, introducing new prompt engineering features like chain-of-thought reasoning and prompt generators that significantly reduce development time, exemplified by their customer Zoominfo. OpenAI teased a "magic" new development coming soon, speculated to be a new LLM replacing GPT-3.5 in the free tier or a search competitor. The open-source community highlighted Llama 3 70B as "game changing" with new quantized weights for Llama 3 120B and CUDA graph support for llama.cpp improving GPU performance. Neuralink demonstrated a thought-controlled mouse, sparking interest in modeling consciousness from brain signals. The ICLR 2024 conference is being held in Asia for the first time, generating excitement.
The Era of 1-bit LLMs
bitnet-b1.58 hugging-face quantization model-optimization energy-efficiency fine-tuning robotics multimodality ai-security ethics humor swyx levelsio gdb npew _akhaliq osanseviero mmitchell_ai deliprao nearcyan clementdelangue
The Era of 1-bit LLMs research, including the BitNet b1.58 model, introduces a ternary parameter approach that matches full-precision Transformer LLMs in performance while drastically reducing energy costs by 38x. This innovation promises new scaling laws and hardware designs optimized for 1-bit LLMs. Discussions on AI Twitter highlight advances in AGI societal impact, robotics with multimodal models, fine-tuning techniques like ResLoRA, and AI security efforts at Hugging Face. Ethical considerations in generative AI and humor within the AI community are also prominent topics.