PhD student | Causality, ML, comp bio @Cambridge_Uni | ex-Applied Scientist Intern @AmazonScience | ex-Visiting researcher @HelmholtzMunich | opinions my own

Joined January 2019
3 Photos and videos
RT @samagreene: In case anyone forgot: Russia went to war in 2014 to prevent Ukraine from signing a Deep and Comprehensive Free Trade Agree…
374
William Orchard retweeted
10 Jul 2025
Long shot but does anyone still need Airbnb for #ICML2025 ? I have booked everything to go but just got diagnosed with an acute disease that now I can't go 😭😭 since not going, the funding body might not reimburse the costs. high appreciate retweeting or contacting re Airbnb 🙏🏻
3
6
8
1,981
William Orchard retweeted
❤️‍🔥 Passionate about gene regulation research, wet/dry lab? 🎡 Excited about London's vibrant research and social scene? 🔬 Looking for a funded postdoc position ( ideally, have the CV & drive to apply for a fellowship in your 1st year)? Please DM/email me to discuss!
4
15
27
6,346
Small clip from of me talking about some work I did at @AmazonScience on Root Cause Analysis! I was really happy to have a chance to briefly chat to @AleksanderMolak about this! (and mercifully he's edited out my terrible rambling - so do please check it out!)
1
11
433
William Orchard retweeted
This is excellent and well worth reading for an informed opinion on the genetics of intelligence. Other kinds of opinion being abundant on this site.
Last week The Atlantic featured an article on the rising popularity of race/IQ science on the right (theatlantic.com/technology/a…). The obvious point that "intelligence is not like height" sparked an unusual amount of whinging. I wrote about how this is now more true than ever. 🧵
1
12
44
9,407
William Orchard retweeted
22 Aug 2024
We will organize a "Causality for LLMs" Tutorial #NeurIPS2024 @NeurIPSConf. Happy to contribute to our community an intro of meaningful topics in Causal LLMs. And super excited for our panel w/ Yoshua, @bschoelkopf @LukeZettlemoyer @_jasonwei @Swarooprm7 @tom4everitt. Stay tuned!
3
55
342
82,928
William Orchard retweeted
iran is announcing a GPT wrapper????
2 Aug 2024
JUST IN: 🇮🇷 Iranian media reports "in coming hours, the world will witness extraordinary scenes and very important developments."
7
60
1,274
32,095
William Orchard retweeted
Monthly I host "data science lunch" in my dept. Today we discussed (among other things) limits on historical inference. This quote from Lewontin came up so I had to track it down (p130 "The evolution of cognition: Questions we will never answer"). Link to PDF in next >
7
22
188
20,432
William Orchard retweeted
27 Jul 2024
when guys say they'd like to have a beer with a candidate everyone claps respectfully but when a woman says she'd like a margarita twenty thousand chuds come out to say women shouldn't vote
94
422
4,160
191,021
William Orchard retweeted
👏Congrats to @MalyValerie for receiving an @alzassociation New to the Field Research Grant to study the role of 3D chromatin in alternative isoform choice in health and dementia!
24 Jul 2024
🙌 Not one, not two, but three VIB researchers received a grant or fellowship from the Alzheimer's Association (@alzassociation)! Congratulations, @DrBilalKhalil, Mohamed Soliman, and @MalyValerie. (@CBD_VIB @CMN_VIB) 🏆 vibbio.tech/3WDJ2v6 🏆
1
11
395
William Orchard retweeted
Due to the high visa cost related to my future PhD, I feel a bit embarrassed to let you know that I've started a gofundme if anyone, no matter sum, would feel comfortable contributing. Would mean the world.💕 gofundme.com/f/eleonoras-dre…
34
175
806
488,844
William Orchard retweeted
Identifying the true structure of the latent data generating process (causal world model) by learning from passive observation alone is impossible without reasonable priors, which transformers don’t have. This is why LLMs are brittle and can’t reliably do counter factual reasoning .
20 Jun 2024
New paper: How can you tell if a transformer has the right world model? We trained a transformer to predict directions for NYC taxi rides. The model was good. It could find shortest paths between new points But had it built a map of NYC? We reconstructed its map and found this:
5
21
197
31,978
William Orchard retweeted
Here's the thing: what current AI (e.g. LLMs) is really doing is memorizing millions of patterns seen in human-generated data, and reapplying them on new inputs. That works great when you're dealing with a well-known problem – until you introduce any amount of novelty.
3
9
176
11,821
William Orchard retweeted
For those who don't subscribe to the New Yorker, I've taken the trouble of screenshotting the entire article. (Sorry, New Yorker! As many people as possible need to read this.) 1/25
51
320
2,235
674,011
I'm in LA for @Conf_CLeaR presenting work I did at Amazon on Root Cause Analysis (arxiv.org/abs/2311.04806)! I loved CLeaR last year and am super excited to be attending again this year. If anyone wants to chat all things RCA, come find me! 🥚🔄🐔
14
359
William Orchard retweeted
We are hiring! Join our lab for an exciting PhD at the intersection of AI, causal inference and neuroscience! Apply here: shorturl.at/jrxCT before January 15th!
6
11
2,958
William Orchard retweeted
Join us in Cambridge! 2x group leader positions. Particular interest: computational biology including AI Core funded. Best start-up package you will get anywhere. jobs.cam.ac.uk/job/43302/

51
80
19,572