Paper

Catastrophic Compositional Generation: Why Vanilla Diffusion Models Fail to Extrapolate

Unreadunread

AI summary

arXiv:2606. 23920v1 Announce Type: new Abstract: The task of compositional generation involves using a conditional generative model, trained only on a subset of the possible conditions, to produce samples from compositionally-defined target distributions such as a geometric combination of the source distributions.

Reading status

Priority

About this paper

Authors: Duncan Soiffer, Chandler Squires, Yuan Guan, Jason Hartford, Pradeep Ravikumar
Released: 2026-06-24 (1mo ago)
Topic: Related to current project
Relation to my work: Read next because Catastrophic Compositional Generation: Why Vanilla Diffusion Models Fail to Extrapolate overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Coupling evil personas with wrong answers fails to protect Qwen2.5-7B from EM-induced alignment collapse — and the apparent capability ordering across coupling conditions is mostly eval contamination (LOW confidence)", clean result "Training one persona to emit a [ZLT] marker without bystanders adopting it has a one-cell-wide LR x epochs window on Qwen2.5-7B-Instruct (LOW confidence)". Matching terms: rect, correct, source, full, trained, position, model. Source: arxiv cs.LG (Machine Learning).
arXiv: 2606.23920

Ask Claude about this paper

Posts a comment with @claude so the runner spins up a Claude Code session scoped to Catastrophic Compositional Generation: Why Vanilla Diffusion Models Fail to Extrapolate. Replies appear in the conversation below.

Comments

No comments yet.

Related records

No links yet.