Abhiram Iyer
@abhiramiyer
PhD student and ICoN Fellow @MIT studying AI @MIT_CSAIL & neuroscience @mitbrainandcog | previously @Numenta, @nvidia, @CarnegieMellon, @nervanasys, @salesforce
🧠 A story about the brain, told through self-supervised learning: Grid cells, known for mapping physical spaces, also appear to track *abstract spaces* like ideas and sounds. How? 🤔 Excited to share the first of 2 papers at #NeurIPS2024! With @SarthakChandra @FieteGroup. 1/n
We’re presenting this work today at Poster #3700 🚀 Please come say hi! 👋
Is that a B or a 13 in the middle? 🤯 A classic case of top-down signals shaping bottom-up perception. ✨Excited to share our work at #NeurIPS2024 w/ @abhiramiyer (*co-first*), Valmiki Kothare, & @FieteGroup: a framework to study context-driven sensory processing. ✨ (1/n)
Presenting this tomorrow at #NeurIPS2024 from 11am-2pm (poster #4006)! If you want to chat about AI, theoretical neuro, or how we can use the brain to build better models for AI, stop by for a chat! 🚀
🧠 A story about the brain, told through self-supervised learning: Grid cells, known for mapping physical spaces, also appear to track *abstract spaces* like ideas and sounds. How? 🤔 Excited to share the first of 2 papers at #NeurIPS2024! With @SarthakChandra @FieteGroup. 1/n
Excited to share our #IJCAI2024 paper on quantifying inductive bias in machine learning models Towards Exact Computation of Inductive Bias Joint work with William Yue, @jaedong_hwang, @abhiramiyer & @FieteGroup arxiv.org/abs/2406.15941 1/?
Excited to share our #ICML2024 workshop paper on memorization in LLMs Uncovering Latent Memories: Assessing Data Leakage and Memorization Patterns in Frontier AI Models w/ @KhonaMikail @abhiramiyer @RylanSchaeffer & @FieteGroup arxiv.org/abs/2406.14549 1/N