[thread] OpenAI o1, o3 | OpenAI GPT-4o
https://en.wikipedia.org/wiki/OpenAI_o1
* generative pre-trained transformer
* form. known within OpenAI as “Q*"
* o1 spends time "thinking" before it answers
* makes it better at complex reasoning tasks, science & programming than OpenAI GPT-4o
* full v. was released 2024-Dec-05
#LLM #OpenAI #OpenAI_o1 #OpenAI_o3 #GPT4o #ML #TransformerArchitecture #reasoning #COT #ChainOfThought #AGI #AI
OpenAI 03 LLM: 87.5% High Score on ARC Prize Challenge
https://old.reddit.com/r/MachineLearning/comments/1hiq3tz/d_openai_o3_875_high_score_on_arc_prize_challenge
https://news.ycombinator.com/item?id=42473321
* GPT-3 scored 0%
* rare benchmark wh. humans get high scores, LLM low scores
* avg. human performance, ARC-AGI is 85%
OpenAI o3 Breakthrough High Score on ARC-AGI-Pub
https://arcprize.org/blog/oai-o3-pub-breakthrough
https://arcprize.org/arc-agi-pub
OpenAI o3 beats 99.8% competitive coders
https://old.reddit.com/r/MachineLearning/comments/1hiqptc/openais_o3_beats_998_competitive_coders_d
#LLM #OpenAI #OpenAI_o1 #OpenAI_o3 #GPT4o #ML #TransformerArchitecture #reasoning #COT #ChainOfThought #AGI #AI
OpenAI’s Next Big AI Effort, GPT-5, Is Behind Schedule and Crazy Expensive
https://www.wsj.com/tech/ai/openai-gpt5-orion-delays-639e7693
nonpaywalled: https://archive.fo/L7fOF
https://news.ycombinator.com/item?id=42485938
* OpenAI ran into problem after problem on its new artificial-intelligence project, code-named Orion
* GPT-5 project, code-named Orion, in works for 18+ mo
* intended as major advancement in tech powering ChatGPT
* limitation: available training data, costs ...
#LLM #AI #AGI #OpenAI #GPT4 #GPT5 #OpenAI_o1 #OpenAI_o3 #ChatGPT
AI quest for human-level intelligence: OpenAI o3 electrifies researchers
https://www.nature.com/articles/d41586-025-00110-6
* record-breaking performance on science/math wows researchers
* OpenAI chatbot model o3 made headlines Dec 2024
* high score on test that marks progress towards artificial general intelligence (AGI)
* OpenAI o3 scored 87.5%, trouncing prev. best score for an artificial intelligence system of55.5%
#LLM #OpenAI #OpenAI_o1 #OpenAI_o3 #GPT4o #ML #AI #transformers #reasoning #COT #ChainOfThought #AGI
Victoria Stuart 🇨🇦 🏳️⚧️