Skip to content
View doerlbh's full-sized avatar
😎
😎

Block or report doerlbh

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
doerlbh/README.md

Baihan Lin

I build learning systems that generalize across domains: foundation models, reinforcement learning, and speech and language decoding, running where the stakes are real.

Currently a professor of AI, psychiatry, and neuroscience at Mount Sinai, and a Senior Computer Scientist with the US Department of Veterans Affairs. Previously Google X / Brain / DeepMind, IBM Research, Amazon, and Microsoft Research.

Books

  • Privacy & Security in Large Language ModelsO'Reilly, 2025
  • Reinforcement Learning Methods in Speech and Language TechnologySpringer, 2024

Selected code

Representations: measuring what a model actually encodes

  • rsatoolbox — the Python library for representational similarity analysis. Contributor; co-author of the eLife 2025 paper.
  • TopologicalRSA — topology and geometry of neural representations. PNAS 2024
  • AGTIC — adaptive geo-topological independence criterion
  • UnsupervisedAttentionMechanism — regularity normalization, a neuroscience-inspired normalization layer. Entropy 2022

Reinforcement learning and bandits

  • BanditZoo — bandits and RL agents in real-world environments. pip install banditzoo. CEC 2022, FUZZ 2022
  • mentalRL — RL agents with mental-disorder propensities. AAMAS 2020
  • dilemmaRL — multi-agent RL in the iterated prisoner's dilemma. PRICAI 2022
  • BerlinUCB — contextual bandits with self-supervision for semi-supervised online learning. AJCAI 2020

Speech, language, and behavior

  • MiniVox — online speaker diarization that learns from scratch, framed as a bandit problem. INTERSPEECH 2020, ACML 2021
  • PsychiatryNLP — NLP over clinical conversation
  • HumanLSTM — predicting human decisions with recurrent networks. IJCNN 2022, PLOS ONE 2022

Elsewhere

  • scTSA — topological simplicial analysis of single-cell RNA-seq. ISMB 2019, Algorithms 2022
  • V2R — real-time augmented reality with projectively realistic webcam projection. IJCAI 2020
  • neuroviz — an interactive visualisation of visual perception in the brain

Reach me

Google Scholar · linlab.org · LinkedIn · doerlbh@gmail.com

Outside of this: music, skateboarding, aikido.

Pinned Loading

  1. rsagroup/rsatoolbox rsagroup/rsatoolbox Public

    Python library for Representational Similarity Analysis

    Python 249 52

  2. UnsupervisedAttentionMechanism UnsupervisedAttentionMechanism Public

    Code for our paper: "Regularity Normalization: Neuroscience-Inspired Unsupervised Attention across Neural Network Layers".

    Jupyter Notebook 20 4

  3. BanditZoo BanditZoo Public

    Python library of bandits and RL agents in different real-world environments

    Python 7 4

  4. mentalRL mentalRL Public

    Code for our AAMAS 2020 paper: "A Story of Two Streams: Reinforcement Learning Models from Human Behavior and Neuropsychiatry".

    Jupyter Notebook 26 7

  5. V2R V2R Public

    Code for our IJCAI 2020 paper: "Keep It Real: a Window to Real Reality in Virtual Reality".

    Python 22 5

  6. MiniVox MiniVox Public

    Code for our ACML and INTERSPEECH papers: "Speaker Diarization as a Fully Online Bandit Learning Problem in MiniVox".

    Cuda 29 5