EdinburghNLP

The Natural Language Processing Group at the University of Edinburgh.

[EdinburghNLP @EdinburghNLP]

Join our PhD programme in Designing Responsible Natural Language Processing at the UKRI AI Centre for Doctoral Training, University of Edinburgh. Applications are now re-opened for Home fee status candidates (past candidates need not re-apply). https://t.co/PkdXiVLEGr

[Pasquale Minervini @PMinervini]

my amazing collaborators will be presenting a few papers at EMNLP/NeurIPS/LoG -- if you're attending any of these conferences, touch base with them; they're great! 🚀🚀🚀
blog post: https://t.co/1qL0LKCPqL

[Ashutosh Adhikari @AshutoshAd63687]

VLMs benefit from tool calls as they allow for "thinking with images".

But can a VLM simply internalise the effects of calling correct tools in the latent space and not need to explicitly call them during inference? 🧵

Presenting tomorrow at 11 AM @COLM_conf !

[Alex Gurung @ EMNLP 🇭🇺 @AlexAag1234]

Did you know training on your own incorrect samples can beat distilling from a much larger teacher?

Traditional RFT only reinforces a narrow set of approaches. Strategically sampling for approach-level diversity gives better pass@k, test-time scaling and RL initialization!
https://t.co/pONJCHEO58
🧵

[Ivan Titov @iatitov]

Very happy to see @j0hngou’s work accepted as a NeurIPS oral! I particularly like the broader idea: let models interact richly during training while pressuring them to reduce that interaction. This gives a denser learning signal and leads to more effective one-shot collaboration

[Pasquale Minervini @PMinervini]

Self-play on a massive number of synthetic environments as an endless source of super-human post-training data 🚀🚀🚀 Really happy this is now out — it may be how we achieve AGI/ASI 🙂

[Pasquale Minervini @PMinervini]

Latent reasoning (i.e. recursively decomposing goals into sub-goals in latent space, and extracting symbolic knowledge by decoding goal and sub-goal representations) waaay before it was cool, straight outta ICML’20

[Pasquale Minervini @PMinervini]

Back in 2021 we managed to get an ICLR Outstanding Paper Award (https://t.co/h1lxuaZFZf) with @_kire_kara_’s amazing MSc thesis work, today this looks a bit more tricky to achieve 😬

[Pasquale Minervini @PMinervini]

Edoardo is amazing, catch this opportunity! 🚀🚀🚀🚀🚀

[Edoardo Ponti @PontiEdoardo]

I am looking for 1 postdoc and 2 fully funded PhDs at Imperial @ICComputing to join my @ERC_Research project AToM ⚛

We aim to build models that learn to compress what they perceive and remember, for permanent memory, longer horizons, and >10x speedups

https://t.co/pa5m2Yvpgk

[Pasquale Minervini @PMinervini]

Fazl and his group are stellar; consider applying!!! 🚀🚀🚀

[Pasquale Minervini @PMinervini]

Whether a model is faithful (and, e.g., whether it's relying on a cue) depends heavily on environmental factors, and faithfulness/mechinterp/safety research should take that into account! 🚀 Amazing work by @aryopg et al.

[Edoardo Ponti @PontiEdoardo]

Developing next-generation accelerators requires benchmarking them on realistic and agentic AI workloads.

With @ARIA_research's Scaling Compute programme, we are releasing TEAS 🍵, serving 5 models (4B to ~1T total) on 9 accelerators for 6 workloads.

https://t.co/5hDR3CR0f5

[Aryo Pradipta Gema @aryopg]

CoT monitoring assumes reasoning traces record what shapes an answer.

Yet, we found that models are less likely to verbalize a cue in its CoT when the cue comes from a tool return instead of the user message! 👀

🧵

[Pasquale Minervini @PMinervini]

Scotland is pretty amazing AI-wise! For example, @EdinburghUni folks have been doing AI research since the 60s (https://t.co/XV2EbAOJq4), and they still have a pretty massive impact (https://t.co/K2v9Ku5QkW, https://t.co/6d3hOBdmeQ, etc.).

[Computational Linguistics Journal @CompLingJournal]

We are pleased to announce the winner of the ACL Computational Linguistics Doctoral Dissertation Award (ACL Dissertation Award)! Congratulations Verna Dankers!
@vernadankers #NLproc #CLJournal

[Pasquale Minervini @PMinervini]

Organisations -- sponsorship opportunities are available for #EACL2027, the flagship European conference in computational linguistics, taking place in Athens in March 2027! Support the #NLP community and connect with researchers and practitioners: https://t.co/PoVIXBtBft

[Verna Dankers @vernadankers]

Today + tomorrow I'm at @aclmeeting '26, and early this morning, I received the 2026 ACL Computational Linguistics Doctoral Dissertation Award!! 🙀🏆👩‍🎓It's the icing on the cake of my excellent PhD experience in Edinburgh @Edin_CDT_NLP @EdinburghNLP. (1/3)

[Ivan Titov @iatitov]

On-policy self-distillation: the teacher saw the solution; the student didn't. Copy that confidence and you gain in-domain but lose out-of-domain reasoning. GeoSD scales the teacher's pull by overlap and bounds drift - OOD +5.7–8.6 pts, ID largely intact. Led by @chatruncata

[Josip Jukic @chatruncata]

Self-distillation gives LLMs dense reasoning feedback without requiring a larger teacher. But ID gains can come at the cost of OOD generalization. New joint work with @iatitov : "Geometric Self-Distillation for Reasoning Generalization" (https://t.co/cfqNwIn3zz)

[Antonio Valerio Miceli Barone @AVMiceliBarone]

Wanna improve your LRM ability to reason logically about code?