Tessera
Folgen

Abhilasha Ravichander

@lasha.bsky.social

Tenure-track faculty at the Max Planck Institute for Software Systems Previously postdoc at UW and AI2, working on Natural Language Processing Recruiting PhD students! 🌐 https://lasharavichander.github.io/

106 Posts4807 Follower1201 Folgt
Abhilasha Ravichander@lasha.bsky.social · 8mo ago

Thrilled to welcome Noam (@dahannoam.bsky.social) as the first intern in our new lab ✨ Noam joined us from the Hebrew University of Jerusalem after five years as a journalist! Looking forward to working together and learning from her unique perspective🤗

Noam Dahan@dahannoam.bsky.social · 8mo ago

Life update: I've joined Max Planck Institute for Software Systems as a research fellow (pre-phd), working with @lasha.bsky.social on factuality and nuanced forms of misinformation. Cheers from Germany!

Maria Antoniak@mariaa.bsky.social · 9mo ago

The project that started my whitespace obsession... #EMNLP2025 While we've all been worrying about tokenizers, lurking in the background has been the preprocessing *before* tokenization. Poems break standard HTML-to-text linearization systems, and we find that multimodal models aren't a solution.

so much depends / upon / a whitespace: Why Whitespace Matters for Poets and LLMs

Whitespace is a critical component of poetic form, reflecting both adherence to standardized forms and rebellion against those forms. Each poem's whitespace distribution reflects the artistic choices ...

arxiv.org

@bhyravajjula.bsky.social · 9mo ago

When you read a poem, do you wonder how the poet structures it through whitespace between/before words and lines? We did! Our findings on whitespace - how to measure/preserve it, how usage varies across form/time, how it affects LLMs - now in an #EMNLP2025 (main) paper: arxiv.org/abs/2510.16713 🧵👇

"so much depends / upon / a whitespace: Why Whitespace Matters for Poets and LLMs" by Sriharsh Bhyravajjula, Melanie Walsh, Anna Preus, Maria Antoniak
Ana Marasović@anamarasovic.bsky.social · 9mo ago

Go check Alex's poster today (Wed) in Suzhou! #EMNLP2025 I'm still so proud of our work (led by @lasha.bsky.social) on CondaQA, so we had to ask what would happen if we tried to create high-quality reasoning-over-text benchmarks now that LLMs are available. Turns out, we'd make an easier benchmark!

Alex Gill@agill32.bsky.social · 9mo ago

I'll be in Suzhou 🇨🇳 at #EMNLP this week presenting "What has been Lost with Synthetic Evaluation?" done with @anamarasovic.bsky.social & @lasha.bsky.social! 🎉 📍Findings Session 1 - Hall C 📅 Wed, November 5, 13:00 - 14:00 arxiv.org/abs/2505.22830

Myra Cheng@myra.bsky.social · 10mo ago

AI always calling your ideas “fantastic” can feel inauthentic, but what are sycophancy’s deeper harms? We find that in the common use case of seeking AI advice on interpersonal situations—specifically conflicts—sycophancy makes people feel more right & less willing to apologize.

Screenshot of paper title: Sycophantic AI Decreases Prosocial Intentions and Promotes Dependence
Maria Ryskina@mryskina.bsky.social · 10mo ago

Interested in language models, brains, and concepts? Check out our COLM 2025 🔦 Spotlight paper! (And if you’re at COLM, come hear about it on Tuesday – sessions Spotlight 2 & Poster 2)!

Paper title: Language models align with brain regions that represent concepts across modalities. Authors: Maria Ryskina, Greta Tuckute, Alexander Fung, Ashley Malkin, Evelina Fedorenko. Affiliations: Maria is affiliated with the Vector Institute for AI, but the work was done at MIT. All other authors are affiliated with MIT. Email address: maria.ryskina@vectorinstitute.ai.
Tiago Pimentel@tpimentel.bsky.social · 10mo ago

LLMs are trained to mimic a “true” distribution—their reducing cross-entropy then confirms they get closer to this target while training. Do similar models approach this target distribution in similar ways, though? 🤔 Not really! Our new paper studies this, finding 4-convergence phases in training 🧵

Figure showing the four phases of convergence in LM training
Abhilasha Ravichander@lasha.bsky.social · 12mo ago

I am recruiting emergency reviewers for *SEM 2025 (The 14th Joint Conference on Lexical and Computational Semantics). Please DM me if you might be able to contribute a review within the next few days 🙏

Maria Antoniak@mariaa.bsky.social · 12mo ago

Status Update: I'm in the middle of my move from Denmark to Colorado! If I seem to be missing for the 1-2 weeks, that is the main reason. Picture me lost amidst suitcases, boxes, moving pods, and far too many books. Copenhagen friends, I'm here for a couple more days! Please stop by P1 to say bye 🥺

Abhilasha Ravichander@lasha.bsky.social · last yr.

Super super thrilled that HALoGEN, our study of LLM hallucinations and their potential origins in training data, received an ✨Outstanding Paper Award✨ at ACL! Joint work w/i Shrusti Ghela*, David Wadden, and Yejin Choi bsky.app/profile/lash...

Abhilasha Ravichander@lasha.bsky.social · 2y ago

We are launching HALoGEN💡, a way to systematically study *when* and *why* LLMs still hallucinate. New work w/ Shrusti Ghela*, David Wadden, and Yejin Choi 💫 📝 Paper: arxiv.org/abs/2501.08292 🚀 Code/Data: github.com/AbhilashaRav... 🌐 Website: halogen-hallucinations.github.io 🧵 [1/n]

Valentina Pyatkin@valentinapy.bsky.social · last yr.

💡Beyond math/code, instruction following with verifiable constraints is suitable to be learned with RLVR. But the set of constraints and verifier functions is limited and most models overfit on IFEval. We introduce IFBench to measure model generalization to unseen constraints.

Bild
Alex Gill@agill32.bsky.social · last yr.

𝐖𝐡𝐚𝐭 𝐇𝐚𝐬 𝐁𝐞𝐞𝐧 𝐋𝐨𝐬𝐭 𝐖𝐢𝐭𝐡 𝐒𝐲𝐧𝐭𝐡𝐞𝐭𝐢𝐜 𝐄𝐯𝐚𝐥𝐮𝐚𝐭𝐢𝐨𝐧? (arxiv.org/abs/2505.22830) I'm happy to announce that the preprint release of my first project is online! Developed with the amazing support of @lasha.bsky.social & @anamarasovic.bsky.social

What Has Been Lost with Synthetic Evaluation?

Large language models (LLMs) are increasingly used for data generation. However, creating evaluation benchmarks raises the bar for this emerging paradigm. Benchmarks must target specific phenomena, pe...

arxiv.org

Nadav Borenstein@nadavb.bsky.social · last yr.

🚨 Preprint alert 🚨 𝐂𝐚𝐧 𝐂𝐨𝐦𝐦𝐮𝐧𝐢𝐭𝐲 𝐍𝐨𝐭𝐞𝐬 𝐑𝐞𝐩𝐥𝐚𝐜𝐞 𝐏𝐫𝐨𝐟𝐞𝐬𝐬𝐢𝐨𝐧𝐚𝐥 𝐅𝐚𝐜𝐭-𝐂𝐡𝐞𝐜𝐤𝐞𝐫𝐬? (arxiv.org/abs/2502.14132) Fact-checking agencies have come under intense scrutiny in recent months regarding their role in combating misinformation on social media.

Can Community Notes Replace Professional Fact-Checkers? Work done by Nadav Borenstein, Greta Warren, Desmond Elliott, and Isabelle Augenstein.