AI Paper Atlas

Research map / Time series / Causal and generative time series

Temporal Island Sparse Autoencoders for Interpreting Clinical Time-Series Models

Muhan Yeo, Bong Gyun Kang, HyunGi Kim, Sungroh Yoon

NeurIPS 2026 · Time series › Causal and generative time series · approach: State-space and latent dynamics forecasting · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Introduces the Temporal Island Sparse Autoencoder (TI-SAE), a backbone-specific sparse autoencoder for interpreting deep clinical time-series models trained on EHR data.

Problem

Deep learning models on electronic health records are accurate but opaque; post-hoc attribution gives per-cell importance scores lacking temporal coherence, cross-variable structure, and reusability across patients.

Approach

A backbone-specific sparse autoencoder with temporal island structure that learns reusable, temporally coherent clinical concept features from a model's internal representations, as an alternative to per-cell attribution scores.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use TI-SAE soft selection instead of integrated gradientsfrozen clinical backbones · MIMIC-III/IV, Discrete backboneAOPC: .060 -> .086Improved
use TI-SAE on STraTS backboneSTraTS · MIMIC-III/IVAOPC: .071 (TSR) vs .016Hurt
add sparse variable routing (ablation)TI-SAE · MIMIC-III/IVAOPC: removal costs 58-75% of AOPCImproved
add temporal islands and channel organization (ablation)TI-SAE · MIMIC-III/IVAOPCNo clear effect
add temporal islands and channel organization (ablation)TI-SAE · MIMIC-III/IVchannel coherence (%): 36.9 -> 91.9Improved
train dictionary before selector instead of applying selection to raw hidden states (ablation)TI-SAE · MIMIC-III/IVAOPC: .038 -> .086 (Discrete)Improved
use multiple islands per concept (M=4) instead of a single island (M=1) (ablation)TI-SAE · MIMIC-III/IVfaithfulness and plausibilityImproved
use learned split-normal window instead of symmetric Gaussian window (ablation)TI-SAE · MIMIC-III/IVfaithfulness and plausibilityNo clear effect
transfer MIMIC-III dictionary to MIMIC-IV without retrainingTI-SAE · MIMIC-IV (2014+)AOPC: .149 -> .130No clear effect
remove dictionary labels from the verbalizer (ablation)TI-SAE · verbalized explanationsLLM judge win rateHurt
use TI-SAE concept explanations instead of IG attributionTI-SAE · verbalized explanations, cliniciansconcept-level reasoning wins: 39-4 (7 ties)Improved
use TI-SAE concept explanations instead of IG attributionTI-SAE · verbalized explanations, cliniciansfaithfulness wins: 4-4 (42 ties)No clear effect

Experimental setup

Baselines: post-hoc attribution methods

Builds on: sparse autoencoders

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy