Picture

Thomas Heap @ThomasEHeap

Bristol ML PhD Student Joined April 2022
  • Tweets
    43
  • Followers
    81
  • Following
    157
  • Likes
    1K
Thomas Heap retweeted
laurence_ai Profile Picture

Laurence Aitchison @laurence_ai

2 weeks ago

Super exciting that my former student @edward_milsom is starting his lab at the University of Bath. Ed has led super-fundamental work on representation learning and learning dynamics. And he’s great fun to work with, so I definitely recommend collaborating!

edward_milsom Profile Picture

Edward Milsom @edward_milsom

2 weeks ago

Super exciting that my former student @edward_milsom is starting his lab at the University of Bath. Ed has led super-fundamental work on representation learning and learning dynamics. And he’s great fun to work with, so I definitely recommend collaborating!

Thomas Heap retweeted
edward_milsom Profile Picture

Edward Milsom @edward_milsom

2 weeks ago

Excited to announce I'll be starting this September 2025 as a Lecturer (Assistant Professor) at the University of Bath! I will continue my research on deep learning foundations, and am open to ideas for collaborations. (Pictured: Bath. Not pictured: University of Bath)

Thomas Heap retweeted
xidulu Profile Picture

Xidulu @xidulu

4 months ago

I talked to a lot of people about "a weight decay paper from Wang and Aitchison" at ICLR, which is officially been accepted at #ICML2025 . Laurence summarized the stuff in our paper in the post, here I will talk about the connection with a *broad* collection of existing works 1/

laurence_ai Profile Picture

Laurence Aitchison @laurence_ai

a year ago

I talked to a lot of people about "a weight decay paper from Wang and Aitchison" at ICLR, which is officially been accepted at #ICML2025 . Laurence summarized the stuff in our paper in the post, here I will talk about the connection with a *broad* collection of existing works 1/

Thomas Heap retweeted
sambowyer__ Profile Picture

Sam Bowyer @sambowyer__

4 months ago

Our position paper on LLM eval error bars has just been accepted to ICML 2025 as a spotlight poster!

sambowyer__ Profile Picture

Sam Bowyer @sambowyer__

6 months ago

Our position paper on LLM eval error bars has just been accepted to ICML 2025 as a spotlight poster!

Thomas Heap retweeted
tslwn Profile Picture

Tim Lawson @tslwn

4 months ago

Here we are! #519 is in Hall 2B, opposite the the D E Shaw & Co stand.

Thomas Heap retweeted
tslwn Profile Picture

Tim Lawson @tslwn

4 months ago

I'm at #ICLR, presenting our work on multi-layer SAEs for language-model interpretability tomorrow (Sat 26 Apr) from 10AM at Hall 3 + Hall 2B #519: iclr.cc/virtual/2025/p…

Thomas Heap retweeted
xidulu Profile Picture

Xidulu @xidulu

5 months ago

#ICLR2025 I will hold two talks on KBLaM (my internship project at MSR Cambridge w/ @jameshensman) at Microsoft’s booth: Thursday and Saturday 4 - 4:30, As well as a poster at Poster Session 5 on Saturday morning.!

MSFTResearch Profile Picture

Microsoft Research @MSFTResearch

6 months ago

#ICLR2025 I will hold two talks on KBLaM (my internship project at MSR Cambridge w/ @jameshensman) at Microsoft’s booth: Thursday and Saturday 4 - 4:30, As well as a poster at Poster Session 5 on Saturday morning.!

Thomas Heap retweeted
tslwn Profile Picture

Tim Lawson @tslwn

5 months ago

Second, we trained SAEs on transformers with randomized parameters, finding that auto-interpretability scores do not always distinguish them from trained models. This underscores the difficulty of automating feature interpretation and the importance of appropriate baselines! 3/

Thomas Heap retweeted
tslwn Profile Picture

Tim Lawson @tslwn

5 months ago

There's a lot to process here, but I was pleased to see that Anthropic's 'Circuit Tracing' paper cites three of our recent contributions to the interpretability literature! 1/

AnthropicAI Profile Picture

Anthropic @AnthropicAI

5 months ago

There's a lot to process here, but I was pleased to see that Anthropic's 'Circuit Tracing' paper cites three of our recent contributions to the interpretability literature! 1/

Thomas Heap retweeted
sambowyer__ Profile Picture

Sam Bowyer @sambowyer__

6 months ago

Really happy to have this paper out on arXiv! Scalable GPU-based Bayesian inference for hierarchical models without requiring gradients wrt model parameters (unlike e.g. VI). arxiv.org/abs/2503.08264

ThomasEHeap Profile Picture

Thomas Heap @ThomasEHeap

6 months ago

Really happy to have this paper out on arXiv! Scalable GPU-based Bayesian inference for hierarchical models without requiring gradients wrt model parameters (unlike e.g. VI). arxiv.org/abs/2503.08264

Thomas Heap retweeted
sambowyer__ Profile Picture

Sam Bowyer @sambowyer__

6 months ago

Our paper on the best way to add error bars to LLM evals is on arXiv! TL;DR: Avoid the Central Limit Theorem -- there are better, simple Bayesian (and frequentist!) methods you should be using instead. Super lightweight library: github.com/sambowyer/baye… 🧵👇

Thomas Heap retweeted
lucyfarnik Profile Picture

Lucy Farnik @lucyfarnik

6 months ago

🚨NEW PAPER ALERT 🚨 SAEs can give us insight into the representations of LLMs. But what about the LLMs' computations? If we want to understand LLMs, we don't just need sparse SAE activations, but also a sparse computational graph connecting them. So how do we get them? A 🧵

Thomas Heap retweeted
edward_milsom Profile Picture

Edward Milsom @edward_milsom

6 months ago

Our paper "Function-Space Learning Rates" is on arXiv! We give an efficient way to estimate the magnitude of changes to NN outputs caused by a particular weight update. We analyse optimiser dynamics in function space, and enable hyperparameter transfer with our scheme FLeRM! 🧵👇

Trends for United States

Dodgers
47 B posts
Michigan
118 B posts
$PACMAN
9.116 posts
Yamamoto
19,5 B posts
Underwood
9.809 posts
UCLA
7.339 posts
Rand
23,5 B posts
Mississippi State
8.248 posts
UNLV
4.188 posts
Oklahoma
31,2 B posts
John Mateer
5.628 posts
#TacosVsSushi
3.777 posts
Orioles
27,5 B posts
#HailState
4.866 posts
Sherrone Moore
2.133 posts
Arizona State
5.819 posts
Tanner Scott
2.981 posts
Aidan Chiles
1.577 posts
Jamo
2.273 posts
Vandy
3.607 posts

You might like

Associate professor at @ucdavis #machinelearning #deeplearning #probability #statistics #optimization #sampling
Krishna Balasubramani...
@krizna_b
548 Followers
Postdoc @ProbAIHub, hosted @warwickstats @Princeton @orfe. PhD @GatsbyUCL. Works on universality and ML theory. He/him. 🌈
Kevin Han Huang
@KevinHanHuang1
331 Followers
Intern @MSFTResearch. PhD student @DTUtweet. Probabilistic ML 🧠 diffusion and sampling🧠. previously intern @SonyAI_global, visitor @NYU_Courant.
Anshuk Uppal
@sigmabayesian
278 Followers
LLMs and probabilistic machine learning. University of Bristol
Laurence Aitchison
@laurence_ai
1K Followers
Finishing PhD student at Imperial College London, working on Machine Learning for Healthcare
Alex Capstick
@alex_capstick_
28 Followers
Research Scientist at @Apple for uncertainty quantification.
Michael Kirchhof
@mkirchhof_
1K Followers
doing a phd in RL/online learning on questions related to exploration and adaptivity
Antoine Moulin
@antoine_mln
1K Followers
Xi Wang, Full-stack Bayesian, ECNU, UMass CICS, JHU CS, Fan of U-Shape
Xidulu
@xidulu
342 Followers
Ph.D. @Stanford and @uwcse
Etash Guha
@etash_guha
648 Followers
AI Scientist @MistralAI | Prev: @AmazonScience @Spotify @Huawei @BristolUni @imperialcollege @UniofOxford
Adam Yang
@adam_x_yang
193 Followers
Taylor
@tbrown122387
71 Followers
Machine Learning: Deep NN, Bayesian Opt.
N. Loka
@nasloka
66 Followers
Interested in Bayesian nonparametrics & Bandits!!!! Prev : Research Engineer @nist, @umdcs
Kalyan Kanagala
@bayesic_bandit
70 Followers
SaritaOlovsky
@OlovskySarita
23 Followers
Researcher in TSE, Toulouse, France. Optimization in ML, Second-Order and Quasi-Newton methods. https://t.co/u56JAxeakv Coffee fan
Dmitry Kamzolov
@KamzolovDmitry
80 Followers