212 Photos and videos
May 26
We evaluated CoT faithfulness evaluations & released 𝐁𝐨𝐧𝐚𝐅𝐢𝐝𝐞 so you can test yours too!!
May 26
Can we tell when LLMs are being unfaithful in their chains of thought?
We evaluated 8 methods claiming to do this, and found that most perform near chance!
But evaluating this requires us to have ground-truth labels for CoT faithfulness. How can we obtain these?
5
16
2,372
Ana Marasović retweeted
May 26
Can we tell when LLMs are being unfaithful in their chains of thought?
We evaluated 8 methods claiming to do this, and found that most perform near chance!
But evaluating this requires us to have ground-truth labels for CoT faithfulness. How can we obtain these?
4
25
147
14,713
Due May 6! NAACL RAF is accepting proposals for 2026-2027! Grants are available for NLP/CompLing initiatives across the Americas. forms.gle/XPxN3MuSVo7rnma57
3
6
1,089
Ana Marasović retweeted
Mar 18
First paper of my PhD at the University of Utah, with Prof @Kenneth_Marino.
Super excited to finally share what we've been working on at SparkLab. Meet TimeWarp ⏳, a benchmark that tests web agents by sending them back in time through 6 eras of web UI design.
Thread 🧵(1/5)
1
2
10
2,388
Ana Marasović retweeted
It’s been less than a year since I started my lab (SPARK Lab) at @UUtah we already have a ton of new stuff that I can’t wait to talk about soon. Stay tuned for more.
I’ll start today by sharing that our updated Computer Use Survey blog has been accepted to ICLR Blogposts 2026.
Collaboration with my student @aplycaebous and Utah colleague @anmarasovic.
1
4
11
1,014
Ana Marasović retweeted
Wanted to share with the CU community that our updated Computer Use Survey blog has been accepted to ICLR Blogposts 2026.
Collaboration with my student @aplycaebousand Utah colleague @anmarasovic.
1
2
6
1,247
Happy new year #NAACL! The 2026 election results are here. Congrats🥳
Chair: Anna Rumshisky @arumshisky
Secretary: Jessy Li @jessyjli
Board members: Muhao Chen @muhao_chen, Francisco (Paco) Guzmán, Ana Marasović @anmarasovic
naacl.org/posts/2025-12-28-N…
Thank you all for voting!
4
34
5,189
Ana Marasović retweeted
10 Dec 2025
💫 On the heels of announcing 12 new faculty fellows last week, SCI's One-U Responsible AI Initiative is excited to add 3 new postdoctoral fellows to its team next year: rai.utah.edu/postdocs-dec-20…
1
2
490
Ana Marasović retweeted
7 Dec 2025
$1,000,000 to understand how LLMs write code.
Announcing: The Martian Interpretability Challenge.
Understanding the inner workings of LLMs is the greatest scientific challenge of our age,. Let's solve it.
Apply here: withmartian.com/prize
🧵👇
11
44
157
31,773
Ana Marasović retweeted
19 Nov 2025
Going through statements from NAACL board candidates. Resonating a lot with many statements, and especially love the one from @anmarasovic!
naacl.org/elections/2026/ind…
1
4
38
4,513
11 Nov 2025
2
288
7 Nov 2025
𝙒𝙚'𝙧𝙚 𝙝𝙞𝙧𝙞𝙣𝙜 𝙣𝙚𝙬 𝙛𝙖𝙘𝙪𝙡𝙩𝙮 𝙢𝙚𝙢𝙗𝙚𝙧𝙨!
Links below because twitter/x is weird.
2
21
88
24,353
7 Nov 2025
Thrilled to see this work recognized at #EMNLP2025!
This framework and approach to measuring CoT faithfulness have been hugely influential for how I think about reasoning evaluation, and I'm so lucky to have worked with such brilliant collaborators. Huge credit to @mtutek
7 Nov 2025
Very honored to be one out of seven outstanding papers at this years' EMNLP :)
Huge thanks to my amazing collaborators @fatemehc__ @anmarasovic @boknilev, this would not have been possible without them!
8
5
64
7,594
Ana Marasović retweeted
6 Nov 2025
Here Comes Another Bubble (AI Edition)
113
478
3,693
425,937
Ana Marasović retweeted
7 Nov 2025
Very honored to be one out of seven outstanding papers at this years' EMNLP :)
Huge thanks to my amazing collaborators @fatemehc__ @anmarasovic @boknilev, this would not have been possible without them!
9
13
115
15,093
4 Nov 2025
Check out Martin's talk at #EMNLP2025 today (Wed)!
If you care about CoT faithfulness, you 𝘮𝘶𝘴𝘵 read this paper. It introduces the first method for measuring CoT faithfulness that is not purely behavioral, but operates with the internals!
31 Oct 2025
Flying out to @emnlpmeeting soon🇨🇳
I'll present our parametric CoT faithfulness work (arxiv.org/abs/2502.14829) on Wednesday at the second Interpretability session, 16:30-18:00 local time A104-105
If you're in Suzhou, reach out to talk all things reasoning :)
1
13
847