diff --git a/_includes/01_research.html b/_includes/01_research.html index af07b5e..c2c3fc8 100755 --- a/_includes/01_research.html +++ b/_includes/01_research.html @@ -21,19 +21,24 @@

Research

are interested in interning at MSR, feel free to reach out over email :)
🔎 - Interpretability. I'm interested in rethinking - interpretability in the context of LLMs + Interpretability methods, especially LLM + interpretability.

augmented imodels - use LLMs to build a transparent model
+ + + attention steering - mechanistically guide LLMs by + emphasizing specific input + spans
explanation penalization - regularize explanations to align models with prior knowledge
adaptive - wavelet distillation - replace neural nets with simple, performant wavelet models + wavelet distillation - replace neural nets with transparent wavelet models
-
+
- 🧠 Neuroscience. Since joining MSR, I've been focused on leveraging LLM interpretability - to understand how the human brain represents language (using fMRI in collaboration with the Huth lab at UT Austin). + 🧠 Semantic brain mapping, mostly using fMRI responses to language. + + +

- explanation-mediated validation - build and test fMRI + explanation-mediated validation - test fMRI explanations using LLM-generated stimuli
- qa embeddings - build interpretable fMRI encoding models by + qa embeddings - predict fMRI language responses by asking yes/no questions to LLMs
summarize & score explanations - generate natural-language - explanations of fMRI encoding models + explanations of fMRI encoding models
💊 - Healthcare. I'm also actively working on how we can improve clinical decision instruments by using - the information contained across various sources in the medical literature (in collaboration with Aaron Kornblith at UCSF and the MSR Health Futures team). + Clinical decision rules, can we improve them with data? + + + + +

+ greedy tree sums - build accurate, compact tree-based clinical + models
clinical self-verification - self-verification improves performance and interpretability of clinical information extraction
clinical rule - vetting - stress testing a clinical decision instrument performance for intra-abdominal injury - + vetting - stress testing a clinical decision instrument performance for intra-abdominal injury
- Across these areas, I'm interested in decision trees and how we can build flexible but accurate transparent - models. I put a lot of my code into the imodels and imodelsX packages.
+ + + Note: I put a lot of my code into the imodels and imodelsX packages. +
@@ -153,6 +164,18 @@

Research

+ + '25 + Vector-ICL: In-context Learning with Continuous Vector Representations + + zhuang et al. + 🔎🌀 + iclr + + + + + '24 Interpretable Language Modeling via Induction-head Ngram Models @@ -175,6 +198,8 @@

Research

+ @@ -187,6 +212,8 @@

Research

+ @@ -200,18 +227,6 @@

Research

- - '24 - Vector-ICL: In-context Learning with Continuous Vector Representations - - zhuang et al. - 🔎🌀 - arxiv - - - - - '24 Towards Consistent Natural-Language Explanations via Explanation-Consistency Finetuning diff --git a/_notes/research_ovws/ovw_interp.md b/_notes/research_ovws/ovw_interp.md index 11af4f7..2f11653 100755 --- a/_notes/research_ovws/ovw_interp.md +++ b/_notes/research_ovws/ovw_interp.md @@ -309,6 +309,12 @@ For an implementation of many of these models, see the python [imodels package]( - longitudinal data, survival curves - misc + + - On the Power of Decision Trees in Auto-Regressive Language Modeling ([gan, galanti, poggio, malach, 2024](https://arxiv.org/pdf/2409.19150)) + - get token word embeddings + - compute exp. weighted avg of embeddings (upweights most recent tokens) + - predicts next embedding with XGBoost (regression loss) then finds closest token + - counterfactuals - [Counterfactual Explanations for Oblique Decision Trees: Exact, Efficient Algorithms](https://arxiv.org/abs/2103.01096) (2021) - [Optimal Counterfactual Explanations in Tree Ensembles](https://arxiv.org/abs/2106.06631) @@ -322,7 +328,7 @@ For an implementation of many of these models, see the python [imodels package]( 1. feature-level: monotonicity, attribute costs, hierarchy/interaction, fairness, privacy 2. structure-level - e.g. minimize #nodes 3. instance-level - must (cannot) link, robust predictions - + - Analysis of Boolean functions ([wiki](https://en.wikipedia.org/wiki/Analysis_of_Boolean_functions)) - Every real-valued function $f:\{-1,1\}^n \rightarrow \mathbb{R}$ has a unique expansion as a multilinear polynomial: diff --git a/_notes/research_ovws/ovw_llms.md b/_notes/research_ovws/ovw_llms.md index ff3ddf8..3933cc4 100644 --- a/_notes/research_ovws/ovw_llms.md +++ b/_notes/research_ovws/ovw_llms.md @@ -1283,7 +1283,7 @@ mixture of experts models have become popular because of the need for (1) fast s - Ravel: Evaluating Interpretability Methods on Disentangling Language Model Representations ([huang, wu, potts, geva, & geiger, 2024](https://arxiv.org/pdf/2402.17700v1.pdf)) -## directly learning algorithms / in-context +## directly learning algorithms - Empirical results - FunSearch: Mathematical discoveries from program search with LLMs ([deepmind, 2023](https://www.nature.com/articles/s41586-023-06924-6)) @@ -1294,6 +1294,10 @@ mixture of experts models have become popular because of the need for (1) fast s - Alphafold - Accurate proteome-wide missense variant effect prediction with AlphaMissense ([deepmind, 2023](https://www.science.org/doi/full/10.1126/science.adg7492)) - predict effects of varying single-amino acid changes - Bridging the Human-AI Knowledge Gap: Concept Discovery and Transfer in AlphaZero ([schut...hessabis, paquet, & been kim, 2023](https://arxiv.org/abs/2310.16410)) +- Learning a Decision Tree Algorithm with Transformers ([zhuang...gao, 2024](https://arxiv.org/abs/2402.03774)) + +## in-context learning + - What Can Transformers Learn In-Context? A Case Study of Simple Function Classes ([garg, tsipras, liang, & valiant, 2022](https://arxiv.org/abs/2208.01066)) - models can succesfully metalearn functions like OLS - e.g. during training, learn inputs-outputs from different linear functions - during testing, have to predict outputs for inputs from a different linear function @@ -1328,6 +1332,12 @@ mixture of experts models have become popular because of the need for (1) fast s - Transformers are Universal In-context Learners ([furuya...peyre, 2024](https://arxiv.org/abs/2408.01367)) - mathetmatically show that transformers are universal and can approximate continuous in-context mappings to arbitrary precision - Limitations - Faith and Fate: Limits of Transformers on Compositionality ([dziri...choi, 2023](https://arxiv.org/abs/2305.18654)) - LLMs can't (easily) be trained well for multiplication (and similar tasks) +- ICLR: In-Context Learning of Representations ([park...wattenberg, tanaka, 2024](https://arxiv.org/abs/2501.00070)) - showing pairs of words sampled from a graph can make the embeddings of those words match the structure of that graph +- Label Words are Anchors: An Information Flow Perspective for + Understanding In-Context Learning ([wang...sun, 2023](https://aclanthology.org/2023.emnlp-main.609.pdf)) +- Correlation and Navigation in the Vocabulary Key Representation Space of Language Models ([peng...shang, 2024](https://arxiv.org/abs/2410.02284)) - some tokens are correlated in embedding space and wrong next-token completions can be highly ranked if their embeddings are correlated with correct ones + - as we sample tokens in context, we get more diverse completions, skipping nearby wrong next tokens + ## cool tasks @@ -1483,7 +1493,8 @@ mixture of experts models have become popular because of the need for (1) fast s - TaBERT: Pretraining for Joint Understanding of Textual and Tabular Data ([yin, neubig, ..., riedel, 2020](https://www.semanticscholar.org/paper/TaBERT%3A-Pretraining-for-Joint-Understanding-of-and-Yin-Neubig/a5b1d1cab073cb746a990b37d42dc7b67763f881)) - classification / predictions - - TabPFN: A Transformer That Solves Small Tabular Classification Problems in a Second ([hollman, ..., hutter, 2022](https://arxiv.org/abs/2207.01848)) + - TabPFN v2: Accurate predictions on small data with a tabular foundation model ([hollman....hutter, 2025](https://www.nature.com/articles/s41586-024-08328-6)) + - TabPFN v1: A Transformer That Solves Small Tabular Classification Problems in a Second ([hollman, ..., hutter, 2022](https://arxiv.org/abs/2207.01848)) - transformer takes in train + test dataset then outputs predictions - each row (data example) is treated as a token and test points attend only to training t - takes fixed-size 100 columns, with zero-padded columns at the end (during training, randomly subsample columns) @@ -1494,7 +1505,7 @@ mixture of experts models have become popular because of the need for (1) fast s - Language models are weak learners ([manikandan, jian, & kolter, 2023](https://arxiv.org/abs/2306.14101)) - use prompted LLMs as weak learners in boosting algorithm for tabular data - TabRet: Pre-training Transformer-based Tabular Models for Unseen Columns ([onishi...hayashi, 2023](https://arxiv.org/abs/2303.15747)) - AnyPredict: A Universal Tabular Prediction System Based on LLMs https://openreview.net/forum?id=icuV4s8f2c - converting tabular data into machine-understandable prompts and fine-tuning LLMs to perform accurate predictions - + - interpretability - InterpreTabNet: Enhancing Interpretability of Tabular Data Using Deep Generative Models and LLM ([si...krishnan, 2023](https://openreview.net/pdf?id=kzR5Cj5blw)) - make attention sparse and describe it with GPT4 @@ -1519,6 +1530,11 @@ mixture of experts models have become popular because of the need for (1) fast s - Embeddings for Tabular Data: A Survey ([singh & bedathur, 2023](https://arxiv.org/abs/2302.11777)) - Deep neural networks and tabular data: A survey ([borisov et al. 2022]()) - mostly compares performance on standard tasks (e.g. classification) +## education + +- Towards Responsible Development of Generative AI for Education: An Evaluation-Driven Approach ([jurenka…ibrahim, 2024](https://storage.googleapis.com/deepmind-media/LearnLM/LearnLM_paper.pdf)) + - seven diverse educational benchmark +- The FACTS Grounding Leaderboard: Benchmarking LLMs' Ability to Ground Response?s to Long-Form Input ([jacovi…das, 2025](https://arxiv.org/abs/2501.03200)) - benchmark evaluates whether responses are consistent with a provided document as context ## llm limitations / perspectives