Evan Hernandez

34 posts

Evan Hernandez banner
Evan Hernandez

Evan Hernandez

@evanqed

ph.d. @mit | building @evidenceopen (we're hiring!) | formerly @google @uwmadison | #nlproc and loud music

Cambridge, MA Katılım Mart 2020
151 Takip Edilen428 Takipçiler
Evan Hernandez
Evan Hernandez@evanqed·
It's been awesome to see this work come together - an agentic interpretability system that iteratively experiments on a model to understand how it works! I believe "LMs with tools" is a powerful paradigm for auto interp. Lots of room for new tools + LMs to increase the scope!
Tamar Rott Shaham@TamarRottShaham

Accepted to #ICML2024!! 🚀 Meet MAIA- a Multimodal Automated Interpretabilty Agent that helps users understand AI systems. Given a user query (eg "label a model’s feature"), MAIA designs experiments iteratively by forming and updating hypotheses based on experimental results.

English
0
5
12
2.4K
Evan Hernandez retweetledi
Sarah Schwettmann
Sarah Schwettmann@cogconfluence·
Thanks @_akhaliq! MAIA is next in our line of work on Interpretability Agents that interrogate other models, and the functions of their components, using iterative experimentation. Project page: …imodal-interpretability.csail.mit.edu/maia/ w/ @TamarRottShaham @f_x_wang @AchyutaBot @evanqed @jacobandreas @ antoniotorralba
AK@_akhaliq

A Multimodal Automated Interpretability Agent This paper describes MAIA, a Multimodal Automated Interpretability Agent. MAIA is a system that uses neural models to automate neural model understanding tasks like feature interpretation and failure mode discovery. It equips a

English
1
15
66
22.5K
Evan Hernandez
Evan Hernandez@evanqed·
Bigger picture, REMEDI is a first step towards methods that can detect anomalies in LM generations (hallucinations, false assertions, etc) before the model even finishes generating. I think these tools have the potential to be valuable guardrails for safely deploying LMs!
English
0
3
15
884
Evan Hernandez
Evan Hernandez@evanqed·
In fact, those directions also do an okay job of detecting errors in the LM’s background knowledge-- which suggests LMs could naturally represent state info and factual info in similar ways.
English
1
0
5
567
Evan Hernandez retweetledi
Jacob Andreas
Jacob Andreas@jacobandreas·
Exciting to see @OpenAI scaling up @evanqed & co's work on learned models for neuron labeling (milan.csail.mit.edu)! One technical observation, and one practical one:
English
3
15
99
22.4K
Evan Hernandez retweetledi
David Bau
David Bau@davidbau·
The NSF has invited us to propose an $18m (RI-1) project for infrastructure to help **you** (academic researchers) to study large language models. What should we build to help your LLM research? Respond with your thoughts. (Or answer 3Q's on a form forms.gle/s7EDW4V12mR7U3…) 🧵
English
7
55
174
43K
Evan Hernandez
Evan Hernandez@evanqed·
Proud of Eric + team for pushing this work! You don't need a giant LM for everything. Relatively small (< 1B param), specially pretrained language models are effective for processing specialized text like EHR notes. Can beat ICL+big models with only O(100) FT data points
Eric Lehman@lehmer16

Recent results have suggested that LLMs encode a surprising amount of clinical knowledge. These results raise an important question about the role of smaller specialized clinical models: do we still need clinical LMs? We explore this question in our paper: arxiv.org/abs/2302.08091

English
0
0
7
767
Evan Hernandez
Evan Hernandez@evanqed·
If you’re attending ICLR this week, come say hi! This work has been a long time coming and we’re excited to share it: Oral: 4/28 4:15AM EST (yikes) Poster: 4/28 9:30PM EST
English
0
0
2
0
Evan Hernandez
Evan Hernandez@evanqed·
Ultimately, MILAN is another tool for our interpretability toolkit. The irony is not lost on us that it relies on a black-box model to understand black-box models. But we think it serves an important purpose in helping practitioners make decisions about their own models.
English
1
0
6
0

Keşfet

@_akhaliq @TamarRottShaham @f_x_wang @AchyutaBot @jacobandreas @belindazli @OpenAI @cogconfluence