Linus Pin-Jie Lin
@linusdd44804
Followers
73
Following
709
Media
18
Statuses
123
PhD @VT_CS, Master @LstSaar. Interested in efficient model development & modular LMs
Saarbrücken
Joined April 2019
Drop by today if you’re around!
I am not at EMNLP this year, but my student @linusdd44804 will be presenting our paper on efficient model development through fine-tuning transfer. The presentation is tomorrow 2-3:30 pm, A109 (session 15). Please come talk to him!
0
0
0
@therealthapa One more @SanghaniCtrVT paper at #EMNLP2025: Efficient Model Development through Fine-tuning Transfer Main proceedings @linusdd44804 @Sub_RBala @tuvllms (all VT) w/@fyliufengyuan, @kandpal_nikhil
https://t.co/OGQa84ots4
aclanthology.org
Pin-Jie Lin, Rishab Balasubramanian, Fengyuan Liu, Nikhil Kandpal, Tu Vu. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing. 2025.
0
3
2
I’ll be presenting our fine-tuning transfer paper tomorrow! TLDR: Alignment tuning effects can be captured as transferable model diff vectors — no need to fine-tune from scratch for every new base model version. Come find me: 🕑 14:00–15:30 📍 A109 (Session 15) #EMNLP2025
Excited to share that our paper on efficient model development has been accepted to #EMNLP2025 Main conference @emnlpmeeting. Congratulations to my students @linusdd44804 and @Sub_RBala on their first PhD paper! 🎉
0
0
5
LoRA makes fine-tuning more accessible, but it's unclear how it compares to full fine-tuning. We find that the performance often matches closely---more often than you might expect. In our latest Connectionism post, we share our experimental results and recommendations for LoRA.
82
564
3K
DeepSeek achieved a strong result on SEAL0, a challenging benchmark for reasoning with conflicting search results. 🎊
Tools & Agents Upgrades 🧰 📈 Better results on SWE / Terminal-Bench 🔍 Stronger multi-step reasoning for complex search tasks ⚡️ Big gains in thinking efficiency 3/5
0
1
5
🎉🎉
Excited to share that our paper on efficient model development has been accepted to #EMNLP2025 Main conference @emnlpmeeting. Congratulations to my students @linusdd44804 and @Sub_RBala on their first PhD paper! 🎉
0
0
1
This work got accepted at Transactions on Machine Learning Research (TMLR). Congratulations to @prateeky2806 and my co-authors. Also, thank you to the reviewers and editors for their time.
Ever wondered if model merging works at scale? Maybe the benefits wear off for bigger models? Maybe you considered using model merging for post-training of your large model but not sure if it generalizes well? cc: @GoogleAI @GoogleDeepMind @uncnlp 🧵👇 Excited to announce my
0
4
13
Ever wondered if model merging works at scale? Maybe the benefits wear off for bigger models? Maybe you considered using model merging for post-training of your large model but not sure if it generalizes well? cc: @GoogleAI @GoogleDeepMind @uncnlp 🧵👇 Excited to announce my
6
87
393
Excited to share that our paper on model merging at scale has been accepted to Transactions on Machine Learning Research (TMLR). Huge congrats to my intern @prateeky2806 and our awesome co-authors @_JLai, @alexandraxron, @manaalfar, @mohitban47, and @TsendeeMTS 🎉!!
Ever wondered if model merging works at scale? Maybe the benefits wear off for bigger models? Maybe you considered using model merging for post-training of your large model but not sure if it generalizes well? cc: @GoogleAI @GoogleDeepMind @uncnlp 🧵👇 Excited to announce my
2
21
90
More thinking power at test-time doesn't fix noisy-search problems—SealQA proves it. AI's reasoning capabilities fall flat when web search turns messy, and SealQA quantifies that. SealQA introduces an exceptionally challenging benchmark for search-augmented language models,
3
3
9
✨ New paper ✨ 🚨 Scaling test-time compute can lead to inverse or flattened scaling!! We introduce SealQA, a new challenge benchmark w/ questions that trigger conflicting, ambiguous, or unhelpful web search results. Key takeaways: ➡️ Frontier LLMs struggle on Seal-0 (SealQA’s
4
41
146
Introducing the DeepSeek-R1 Thoughtology -- the most comprehensive study of R1 reasoning chains/thoughts ✨. Probably everything you need to know about R1 thoughts. If we missed something, please let us know.
Models like DeepSeek-R1 🐋 mark a fundamental shift in how LLMs approach complex problems. In our preprint on R1 Thoughtology, we study R1’s reasoning chains across a variety of tasks; investigating its capabilities, limitations, and behaviour. 🔗: https://t.co/Cyy18kYQ45
0
24
83
How does RL improve performance on math reasoning? Studying RL from pretrained models is hard, as behavior depends on choice of base model. 🚨 In our new work, we train models *from scratch* to study the effect of the data mix on the behavior of RL. https://t.co/XtToYfkFiP
3
37
144
🚨 New paper 🚨 Excited to share my first paper w/ my PhD students!! We find that advanced LLM capabilities conferred by instruction or alignment tuning (e.g., SFT, RLHF, DPO, GRPO) can be encoded into model diff vectors (à la task vectors) and transferred across model
14
94
445
Almost 7 years ago, Tu Vu and I wrote our first paper together, one of few. It is fantastic to see the first paper by Tu’s students this time. Congratulations and looking forward to many such great works from Tu’s group!
🚨 New paper 🚨 Excited to share my first paper w/ my PhD students!! We find that advanced LLM capabilities conferred by instruction or alignment tuning (e.g., SFT, RLHF, DPO, GRPO) can be encoded into model diff vectors (à la task vectors) and transferred across model
1
2
15
My first PhD paper is out 😆 took 7 months and lots of back-and-forth. Learned so much from Tu — sharp thinking, real feedback, and always pushing the idea further. Also, shoutout to my collaborators and the folks at @VT_CS!
Our paper is now available on arXiv:
0
0
4
Our paper is now available on arXiv:
arxiv.org
Modern LLMs struggle with efficient updates, as each new pretrained model version requires repeating expensive alignment processes. This challenge also applies to domain- or languagespecific...
🚨 New paper 🚨 Excited to share my first paper w/ my PhD students!! We find that advanced LLM capabilities conferred by instruction or alignment tuning (e.g., SFT, RLHF, DPO, GRPO) can be encoded into model diff vectors (à la task vectors) and transferred across model
0
3
12