---
title: "Cross-Prompt Reuse of Entity-Replacement Responses"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2611.00352v1"
permalink: "https://feed.astralor.com/items/itm_e3bf9d09f2bda01e581cc195"
published: "2026-10-30T20:32:53.000Z"
score: 57
channel: "ai"
tags: ["Cross-Prompt Reuse","Entity-Replacement Responses With","English","With","Hidden","Qwen"]
---

# Cross-Prompt Reuse of Entity-Replacement Responses

## At a glance

With model weights fixed, we measure how much of an entity-replacement response transfers between English factual questions, beyond whether answers change. We add the hidden-state change from two source questions to a t…

## Key points

- With model weights fixed, we measure how much of an entity-replacement response transfers between English factual questions, beyond whether answers change.
- We add the hidden-state change from two source questions to a target question about the same entity, comparing the predicted log-probability shift between two relations' published answers, given known answer prefixes, w…
- Hidden reuse injects the aligned change at 50% depth before the target's remaining layers; direct reuse adds source output-score changes.

## Full text

With model weights fixed, we measure how much of an entity-replacement response transfers between English factual questions, beyond whether answers change. We add the hidden-state change from two source questions to a target question about the same entity, comparing the predicted log-probability shift between two relations' published answers, given known answer prefixes, with the measured one. Hidden reuse injects the aligned change at 50% depth before the target's remaining layers; direct reuse adds source output-score changes. With other-relation sources, hidden reuse removes 45% and 49% of no-change error on 0.5-1B Qwen and Llama. Gemma-2-2B removes 22% and passes one of three comparisons in plain text without its default beginning-of-sequence (BOS) token. A later protocol restoring it yields 58-68%, passing every comparison on three cohorts. Larger checkpoints, chat format and new entities show replication and format- or cohort-dependent failures. On small plain-format checkpoints, other-relation hidden reuse mainly reproduces the shift both relations' questions share, without reliably lowering relation-selective error. With target-relation sources, hidden reuse's post hoc advantage over direct reuse is 0.2-0.9 nats outside chat format, unconfirmed and not replacement-specific. The same relation-neutral change at both questions' entity tokens lowers relation-selective error more than random or other-replacement changes, indicating receiving-question dependence.

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_e3bf9d09f2bda01e581cc195