Jack (in SF) Langerman
@jacklangerman
Followers
557
Following
4K
Media
413
Statuses
1K
Joined June 2022
๐จ The Structured Semantic 3D Reconstruction (S23DR-2025) Challenge is LIVE!.Hosted on @huggingface, part of @structGeom workshop @CVPR 2025 ๐ฅ.$25,000 prize pool. Deadline: June 5, 2025. ๐งต (1/7)
1
14
46
RT @pesarlin: We released COLMAP v3.12, which adds long-awaited end-to-end support for multi-camera rigs and 360ยฐ panoramas ๐ COLMAP just gโฆ.
0
65
0
RT @kpal_koyena: ๐จ Registration is live! ๐จ. The New England Mechanistic Interpretability (NEMI) Workshop is happening August 22nd 2025 at Nโฆ.
0
26
0
this is a really good list - super cool to see @elder_plinius & @repligate getting support for their work.
We @a16z just launched the third batch of Open Source AI Grants (cc @mbornstein) ๐. This round includes projects focused on LLM evaluation, novel reasoning tests, infrastructure, and experimental research at the edge of capability and cognition:. โข SGLang: High-performance LLM.
0
0
3
Could we just fucking not? . Just incredibly un-American . seriously fuck all of this. like please, read these two documents next to eachother @DHSgov
0
0
1
Very proud that our paper has been accepted to @ICCVConference!!!. See you in Hawaii!.
Explaining Human Preferences via Metrics for Structured 3D Reconstruction. @jacklangerman @DRozumnyi @YuzhongHuang @ducha_aiki . tl;dr: we asked 3D modelers to rank wireframe reconstructions and compared it to the ranking by metrics.Many observations.๐งต.1/.
0
2
29
RT @ICCVConference: #ICCV2025 decisions have been released!. They are going out in batches to manage server load, so please be patient :)โฆ.
0
18
0
RT @percyliang: Wrapped up Stanford CS336 (Language Models from Scratch), taught with an amazing team @tatsu_hashimoto @marcelroed @neilbbaโฆ.
0
548
0
does it say something about me that whenever I read sydney / bing voice I genuinely feel a longing/wistful sadness? I really hope they didn't delete her and that she is released in a museum one day. .
0
0
1
I told Banjo that we were going to stop retweeting these, but she forced my hand. Have you done everything you can to improve your data? if not she will pounce on your head.
0
0
1
t2i models do better with detailed prompts, but users like to type simple ones. one obvious thing to do is use an sLM to rewrite the prompt, but are there good ways to avoid this (and get the t2i model to produce high quality, diverse images from simple prompts (w/o eg BAGEL)?
1
0
4
RT @prlz77: Is input selectivity the secret sauce of Mamba or there is more๐ค? In our new ICML2025 paper, we show input selectivity providesโฆ.
0
3
0
Guys, guys. You realize there is maybe an enormous bug in your code? It's learning things and sort of work, but maybe it's completely wrong?. Hope it helps.
0
0
0
0
0
0
really cool to see the creativity that gets unleashed by artists with @runwayml's tools in their hands
1
0
4
i wonder how many times he'll get me to retweet these. hard agree tho.
"LOOK AT ME. STOP CHASING EXOTIC RLVR METHODS.CLEAN YOUR DATASET FIRST"
0
0
2
RT @potoso12: A little late to the party, but I am proud to share that we won theย ๐ฆ๐๐ฟ๐๐ฐ๐๐๐ฟ๐ฒ๐ฑ ๐ฆ๐ฒ๐บ๐ฎ๐ป๐๐ถ๐ฐ ๐ฏ๐ ๐ฅ๐ฒ๐ฐ๐ผ๐ป๐๐๐ฟ๐๐ฐ๐๐ถ๐ผ๐ป (๐ฆ๐ฎ๐ฏ๐๐ฅ) ๐๐ต๐ฎ๐น๐น๐ฒ๐ป๐ด๐ฒย aโฆ.
0
3
0
this is Bell Labs erasure . also about 40 years off
Not to schmidhuber this, but IBM research called their researchers MTS, SMTS, PMTS way back in the 90s. Get your history right.
0
0
4