Bob West
@cervisiarius
Followers
2K
Following
219
Media
17
Statuses
285
Associate Professor at EPFL, Data Science Lab (dlab)
Lausanne, Switzerland
Joined September 2010
RT @Butanium_: This work got accepted to ACL 2025 main! š.In this updated version, we extended our results to several models and showed theā¦.
0
5
0
RT @wendlerch: How do diffusion models create images and can we control that process? . We are excited to release a update to our SDXL Turbā¦.
0
18
0
RT @VminVsky: New paper: Language models have āuniversalā concept representation ā but can they capture cultural nuance? š. If someone fromā¦.
0
33
0
RT @aroraakhilcs: I am recruiting 2 PhD students for Fall'25 @csaudk to work on bleeding-edge topics in #NLProc #LLMs #AIAgents (e.g. LLM rā¦.
0
20
0
New CACM opinion piece written together with Roland Aydin about what we call the āAI Alignment Paradoxā: virtuous AI may be more easily made vicious. Possibly this might be a reason for an intriguing recent result by @OwainEvans_UK et al.:
"The AI Alignment Paradox," by @cervisiarius (@EPFL_en) and Roland Aydin, says that aligning #AIModels with our values makes it easier for adversaries to misalign them.
0
5
11
RT @CACMmag: "The AI Alignment Paradox," by @cervisiarius (@EPFL_en) and Roland Aydin, says that aligning #AIModels with our values makes iā¦.
0
2
0
RT @tzushengkuo: Excited to announce the #WikiNLP workshop at @aclmeeting!. We welcome #NLP contributions to Wikimedia, especially on datasā¦.
0
9
0
If youād like to meet up in Vancouver, get in touch! (Best via email rather than DM on X, which I somehow often miss.).
0
0
2
Looking forward to attending #NeurIPS next week, to catching up with as many old colleagues as possible, and to meeting plenty of future ones!.
1
1
14
New work with @krisgligoric @erichorvitz @emrek @swissepi and Ryen White on food choice mimicry at @EPFL_en.
To nudge people to eat better, it helps to understand what shapes food choices. A study of 18 million purchases at a canteen at @EPFL_en shows that after a person selects an item, the next person in line is more likely to choose the same item. PNAS Nexus:
1
5
12
I love this paper, one of my recent favorites ā not just because of the amazing animated llama in @jkminderās tweet :).
Can we understand and control how language models balance context and prior knowledge? Our latest paper shows itās all about a 1D knob! šļø.Co-led with @kevdududu, as well as @niklas_stoehr, @giomonea, @wendlerch, @cervisiarius & Ryan Cotterell.
0
1
8
RT @jkminder: Can we understand and control how language models balance context and prior knowledge? Our latest paper shows itās all aboutā¦.
0
22
0
RT @aroraakhilcs: Disappointed to not be at #EMNLP owing to a dislocated shoulder š¢.@DebjitPaul2 will present our poster on Multilingual Eā¦.
0
11
0
RT @ViaSurkov: Excited to share our latest breakthrough! We trained sparse autoencoders to decompose intermediate results of SDXL Turbo's fā¦.
0
12
0
RT @russogiusep: Our paper "The AI Review Lottery: Widespread AI-Assisted Peer Reviews Boost Paper Scores and Acceptance Rates" got coveredā¦.
0
5
0
RT @niklas_stoehr: Our new mechanistic interpretability work "Activation Scaling for Steering and Interpreting Language Models" was accepteā¦.
0
18
0
RT @im_td: shout out to my amazing collaborators: @ViaSurkov, @VminVsky, @russogiusep, @cervisiarius, and @caglarml . paper link: https://tā¦.
0
2
0
Exciting 4-yr PhD position pushing the boundaries of LLMs for engineering at the Helmholtz-Zentrum Hereon / TU Hamburg, supervised by the amazing Roland Aydin & co-supervised by me. Applications due on 1 Sep! #PhDPosition #ML #Engineering #LLM #Hamburg.
0
11
35
RT @wendlerch: Very grateful to have received an outstanding paper award for our āDo Llamas Work in English?ā paper. This paper would not hā¦.
0
6
0
RT @krisgligoric: Science faces a reproducibility crisis. Data science & AI courses often include hands-on projects using real-world dataseā¦.
0
19
0