Paper

Anchor PCA

Unreadunread

AI summary

arXiv:2606. 06233v1 Announce Type: new Abstract: Principal component analysis (PCA) is one of the most widely used unsupervised dimension reduction techniques.

Reading status

Priority

About this paper

Authors: Benedikt Seiter, Anya Fries, Julius von K\"ugelgen, Jonas Peters
Released: 2026-06-05 (2mo ago)
Topic: Related to current project
Relation to my work: Read next because Anchor PCA 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 "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)", clean result "A pretraining-data-poisoned Qwen3-4B backdoor only fires on the exact trigger tokens — paraphrases don't activate it, and base-model similarity to the trigger doesn't predict which inputs fire (MODERATE confidence)". Matching terms: rect, under, line, rate. Source: arxiv stat.ML (Machine Learning).
arXiv: 2606.06233

Ask Claude about this paper

Posts a comment with @claude so the runner spins up a Claude Code session scoped to Anchor PCA. Replies appear in the conversation below.

Comments

No comments yet.

Related records

No links yet.