AI Paper Atlas

Research map / Continual learning / Continual fine-tuning of LLMs

One-Prompt Strikes Back: Sparse Mixture of Experts for Prompt-based Continual Learning

Minh Le, Bao-Ngoc Dao, Huy Nguyen, Quyen Tran, Anh Nguyen, Nhat Ho

ICLR 2026 · Continual learning › Continual fine-tuning of LLMs · approach: Low-rank and model-merging adaptation · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use a sparse MoE shared prompt (SMoPE) instead of single shared prompt OVOROVOR · ImageNet-R (10 tasks)FAA: 75.25 -> 79.32Improved
use SMoPE instead of VQ-PromptVQ-Prompt · CUB-200 (10 tasks)FAA: 86.72 -> 87.43Improved
use SMoPE instead of VQ-Prompt on DINO-1KVQ-Prompt · ImageNet-R (DINO-1K)FAA: 68.42 -> 68.61No clear effect
add sparse expert selection to a single shared prompt (ablation)One Prompt (prefix tuning) · ImageNet-RaccuracyImproved
add adaptive noise for balanced expert utilization (ablation)SMoPE · ImageNet-RaccuracyImproved
add prompt-attention score aggregation alone (ablation)One Prompt (prefix tuning) · ImageNet-RaccuracyNo clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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