[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.] [O3](/topic/o3) ### Top Social Posts *Showing only X posts for non-authenticated requests. Use your API key in requests for full results.* "@petergostev @lmarena_ai how do we know it's o3 and not o4-full" [@CtrlAltDwayne](/creator/x/CtrlAltDwayne) on [X](/post/tweet/1946352068723118288) 2025-07-18 23:31:09 UTC 3516 followers, 2579 engagements "Claude and Gemini seem more deferential to citations than o3 leading to different error types -- Claude/Gemini are reasonably opposed to hallucinating info out of nowhere but are insufficiently opposed to parroting false conventional wisdom or just saying nothing useful" [@Miles_Brundage](/creator/x/Miles_Brundage) on [X](/post/tweet/1946340158443499766) 2025-07-18 22:43:49 UTC 59.6K followers, 1837 engagements "@petergostev @lmarena_ai im wondering if this is an o3 update. i dont think its their open-source model. many people complained that o3 performed poorly at general coding compared to other models so i think this is OpenAI improving it before GPT-5 since it will take some time to release" [@thgisorp](/creator/x/thgisorp) on [X](/post/tweet/1946336546421866505) 2025-07-18 22:29:28 UTC XX followers, 2475 engagements "GPT-5 expectations: - SOTA on most benchmarks (#1 on my meta-benchmark) - specifically: SOTA on ARCAGI2 and METR - 2025 knowledge cutoff - longer context window (400k) - fully multimodal (text/image/audio + video input) - sane output pricing: = $XX / 1M tokens nicetohaves: - fewer hallucinations than o3 - less sycophancy - no more barrage of em dashes - clean code style no weird multiline comments" [@scaling01](/creator/x/scaling01) on [X](/post/tweet/1946335109277274344) 2025-07-18 22:23:45 UTC 16K followers, 1157 engagements
[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.]
Showing only X posts for non-authenticated requests. Use your API key in requests for full results.
"@petergostev @lmarena_ai how do we know it's o3 and not o4-full"
@CtrlAltDwayne on X 2025-07-18 23:31:09 UTC 3516 followers, 2579 engagements
"Claude and Gemini seem more deferential to citations than o3 leading to different error types -- Claude/Gemini are reasonably opposed to hallucinating info out of nowhere but are insufficiently opposed to parroting false conventional wisdom or just saying nothing useful"
@Miles_Brundage on X 2025-07-18 22:43:49 UTC 59.6K followers, 1837 engagements
"@petergostev @lmarena_ai im wondering if this is an o3 update. i dont think its their open-source model. many people complained that o3 performed poorly at general coding compared to other models so i think this is OpenAI improving it before GPT-5 since it will take some time to release"
@thgisorp on X 2025-07-18 22:29:28 UTC XX followers, 2475 engagements
"GPT-5 expectations: - SOTA on most benchmarks (#1 on my meta-benchmark) - specifically: SOTA on ARCAGI2 and METR - 2025 knowledge cutoff - longer context window (400k) - fully multimodal (text/image/audio + video input) - sane output pricing: = $XX / 1M tokens nicetohaves: - fewer hallucinations than o3 - less sycophancy - no more barrage of em dashes - clean code style no weird multiline comments"
@scaling01 on X 2025-07-18 22:23:45 UTC 16K followers, 1157 engagements
/topic/o3/posts