AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

The Deleuzian Representation Hypothesis

Clément Cornet, Romaric Besançon, Hervé Le Borgne

ICLR 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Self-explaining and visual explanation networks · 0 citations (OpenAlex)

Read the paper

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
cluster activation differences instead of raw activations (ablation)Deleuzian KMeans · WikiArt artist (CLIP) / CoNLL NER (DeBERTa)probe loss: 0.0133 -> 0.0119 (CLIP); 0.1184 -> 0.0665 (DeBERTa)Improved
replace KMeans with Tk-SAE on activation differences (ablation)Deleuzian · WikiArt artist (CLIP) / CoNLL NER (DeBERTa)probe loss: 0.0119 -> 0.0134 (CLIP); 0.0665 -> 0.1093 (DeBERTa)Hurt
add inverse-skewness weighting to KMeans (ablation)Deleuzian · WikiArt artist (CLIP) / CoNLL NER (DeBERTa)probe loss: 0.0128 -> 0.0119 (CLIP); 0.0841 -> 0.0665 (DeBERTa)Improved
add inverse-skewness weighting to KMeans (ablation)Deleuzian · WikiArt artist (CLIP) / CoNLL NER (DeBERTa)effective rank: 17.9 -> 124.4 (CLIP); 5.65 -> 182.0 (DeBERTa)Improved
add inverse-skewness weighting to KMeans (ablation)Deleuzian · WikiArt artist (CLIP) / CoNLL NER (DeBERTa)max. pairwise cosine: 0.6504 -> 0.5677 (CLIP); 0.8357 -> 0.3908 (DeBERTa)Improved
Deleuzian vs Tk-SAE effective rank on differences (ablation)Deleuzian · WikiArt artist (CLIP) / CoNLL NER (DeBERTa)effective rank: Tk-SAE on diff 340.5 / 109.2 vs Deleuzian 124.4 / 182.0Mixed
use Deleuzian concepts instead of best SAE (Tk-SAE)unsupervised concept extraction · 20 probe-loss tasksprobe loss: lowest probe loss on 13 of 20 tasks; avg rank 1.65 vs 2.65Improved
use Deleuzian concepts instead of Van-SAEconcept extraction · MPPC consistency across seedsMPPC: 0.821 vs 0.840 (CLIP ImageNet); 0.789 vs 0.603 (DinoV2 ImageNet)Mixed
replace linear with quadratic Deleuzian extension (ablation)Deleuzian · WikiArt (CLIP)probe loss: 0.0119 -> 0.0124 (artist); 0.1230 -> 0.1305 (genre)Hurt
use only 2000 concepts instead of 6144 (ablation)Deleuzian · WikiArt artist (CLIP)probe lossNo clear effect
Deleuzian vs supervised LDA on BART CoNLL-2003Deleuzian · CoNLL-2003 NER (BART)probe loss: 0.6326 (LDA) -> 0.2148Improved

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