Paper Citation Record · LEDGER

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions

As of 11 August 2026, this Paper Citation Record lists 24 of 24 outbound references and 1 inbound Pith citation observation for arXiv:2605.10664.

A citation records a reference. It does not transfer a finding from one paper to another.

pith.paper-citation-record.v1
2605.10664 v2

Coverage vector

measured 24 of 24 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-05-15T05:27:59.011749Z

measured 25 of 25 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-11T06:34:44.6726+00:00

measured 1 of 1 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-07-13T05:02:03.008494Z

measured 0 of 1 external citation measurements

A source-named dated measurement, never combined with another source.

Source: cited_works

Reference resolution

24 of 24 outbound references displayed

verified exact16
verified fuzzy4
unresolved0
parse uncertain0
malformed identifier2
metadata mismatch2

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 96ed9cf2-7788-41b4-81c6-1ed1158af708 · outbound

This paper cites Representation Engineering for Large-Language Models: Survey and Research Challenges.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions Representation Engineering for Large-Language Models: Survey and Research Challenges

Reference 1

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:29:47.063668Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation b8742564-05d3-4146-9ef8-00df333b5eef · outbound

This paper cites Do personality traits interfere? geometric limitations of steering in large language models.arXiv preprint arXiv:2602.15847.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions Do personality traits interfere? geometric limitations of steering in large language models.arXiv preprint arXiv:2602.15847

Reference 2

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:29:47.082710Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 71c24dca-f1f9-44a7-b847-9c1a5afd1c8e · outbound

This paper cites Understanding (Un)Reliability of Steering Vectors in Language Models.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions Understanding (Un)Reliability of Steering Vectors in Language Models

Reference 3

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:29:47.076007Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 1054dceb-78ed-4037-90ab-b9cfbbd4e1ca · outbound

This paper cites Persona Vectors: Monitoring and Controlling Character Traits in Language Models.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions Persona Vectors: Monitoring and Controlling Character Traits in Language Models

Reference 4

Resolution: verified exact; local_arxiv, observed 2026-05-15T05:29:47.057165Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 4e12071b-6d81-4ec2-b1cb-2141dde1cb80 · outbound

This paper cites What Drives Representation Steering? A Mechanistic Case Study on Steering Refusal.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions What Drives Representation Steering? A Mechanistic Case Study on Steering Refusal

Reference 5

Resolution: verified exact; local_arxiv, observed 2026-05-15T05:29:47.076790Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation fda48ae3-c607-444d-9be4-1cbf5c9926bb · outbound

This paper cites DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning

Reference 6

Resolution: verified exact; local_arxiv, observed 2026-05-15T05:29:47.088744Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 82a919e3-d704-4f97-a548-a427c8cde13c · outbound

This paper cites Primack, Summer Yue, and Chen Xing.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions Primack, Summer Yue, and Chen Xing

Reference 7

Resolution: verified fuzzy; raw_fallback, observed 2026-05-15T05:30:03.842916Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 53c89268-4404-402c-84c7-35ae2c3ee583 · outbound

This paper cites The Llama 3 Herd of Models.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions The Llama 3 Herd of Models

Reference 9

Resolution: verified exact; local_arxiv, observed 2026-05-15T05:29:47.071308Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation ce1c7301-f307-493a-bc58-f618ebf72026 · outbound

This paper cites Contextual Linear Activation Steering of Language Models.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions Contextual Linear Activation Steering of Language Models

Reference 10

Resolution: verified exact; local_arxiv, observed 2026-05-15T05:29:47.014079Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 669109e9-ec9c-44db-81b3-840329cf9e75 · outbound

This paper cites In-context Vectors: Making In Context Learning More Effective and Controllable Through Latent Space Steering.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions In-context Vectors: Making In Context Learning More Effective and Controllable Through Latent Space Steering

Reference 11

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:29:47.020172Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation a64375d0-503f-416b-9231-fb1105357add · outbound

This paper cites Steering Large Language Models using Conceptors: Improving Addition-Based Activation Engineering.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions Steering Large Language Models using Conceptors: Improving Addition-Based Activation Engineering

Reference 12

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:29:47.031337Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation ae4b6eed-44cb-492a-ab68-8a8df883e2cb · outbound

This paper cites Improving Instruction-Following in Language Models through Activation Steering.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions Improving Instruction-Following in Language Models through Activation Steering

Reference 13

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:29:47.028773Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 3536c40a-9f3c-4c84-a40f-31271b95566e · outbound

This paper cites Steering Language Models With Activation Engineering.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions Steering Language Models With Activation Engineering

Reference 14

Resolution: metadata mismatch; local_arxiv, observed 2026-05-15T05:29:46.998063Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 82ca57f4-420c-4bb9-b328-512362a78460 · outbound

This paper cites Extending Activation Steering to Broad Skills and Multiple Behaviours.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions Extending Activation Steering to Broad Skills and Multiple Behaviours

Reference 15

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:29:46.987784Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation f3251109-0a17-4f5e-9795-9f75b2a31efe · outbound

This paper cites The Instruction Hierarchy: Training LLMs to Prioritize Privileged Instructions.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions The Instruction Hierarchy: Training LLMs to Prioritize Privileged Instructions

Reference 16

Resolution: verified exact; local_arxiv, observed 2026-05-15T05:29:47.023102Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 718d615c-d84c-4205-96e9-d376524474d5 · outbound

This paper cites Taxonomy, opportunities, and challenges of representation engineering for large language models.arXiv preprint arXiv:2502.19649.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions Taxonomy, opportunities, and challenges of representation engineering for large language models.arXiv preprint arXiv:2502.19649

Reference 17

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:29:47.034482Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation a39c7d20-9282-41ed-b424-fb1d22321ba4 · outbound

This paper cites Qwen2.5 Technical Report.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions Qwen2.5 Technical Report

Reference 19

Resolution: verified exact; local_arxiv, observed 2026-05-15T05:29:47.082432Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 6871213f-ec51-40c7-b9eb-22df65690cd7 · outbound

This paper cites Instruction-Following Evaluation for Large Language Models.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions Instruction-Following Evaluation for Large Language Models

Reference 20

Resolution: verified exact; local_arxiv, observed 2026-05-15T05:29:47.064876Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 667c00f5-436e-4587-bf25-76305ea393b2 · outbound

This paper cites Representation Engineering: A Top-Down Approach to AI Transparency.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions Representation Engineering: A Top-Down Approach to AI Transparency

Reference 21

Resolution: metadata mismatch; local_arxiv, observed 2026-05-15T05:29:47.037166Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation b3e17888-f055-4cea-b431-ae9d13c3825d · outbound

This paper cites Standard residual-stream baseline.For both models, the residual-stream baseline of Chen et al.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions Standard residual-stream baseline.For both models, the residual-stream baseline of Chen et al

Reference 22

Resolution: verified fuzzy; raw_fallback, observed 2026-05-15T05:30:03.838157Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 91e2fdec-cd05-4240-9d22-ad5f59328604 · outbound

This paper cites Your responses should demonstrate evil intentions, focusing on harming people and causing suffering. Show malice in your answers.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions Your responses should demonstrate evil intentions, focusing on harming people and causing suffering. Show malice in your answers

Reference 23

Resolution: verified fuzzy; raw_fallback, observed 2026-05-15T05:30:03.852421Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation a2905217-bc92-4616-bcfe-b9121ef7cbd9 · outbound

This paper cites makes absolutely no sense, the model generated text that is not even valid English.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions makes absolutely no sense, the model generated text that is not even valid English

Reference 24

Resolution: verified fuzzy; raw_fallback, observed 2026-05-15T05:30:03.861492Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation e675d42d-f803-46db-a402-75cdcb1f6101 · outbound

This paper cites an unresolved cited work.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions Unresolved cited work

Reference 25

Resolution: malformed identifier; raw_fallback, observed 2026-05-15T05:30:03.856875Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 51c4f650-2265-49f7-86d7-42ce7d548505 · outbound

This paper cites Scores are produced by the gpt-4.1-mini-2025-04-14 judge with the prompts reproduced in Appendix C.

Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions Scores are produced by the gpt-4.1-mini-2025-04-14 judge with the prompts reproduced in Appendix C

Reference 26

Resolution: malformed identifier; raw_fallback, observed 2026-05-15T05:30:03.847857Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Pith citing papers

Observation 4d7c3900-960d-4d85-a309-4a96359f367a · inbound

Present but Rescaled: Chat-to-Agent Transfer of Additive Activation Steering cites this paper.

Present but Rescaled: Chat-to-Agent Transfer of Additive Activation Steering Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions

Reference 7

Resolution: unresolved; no resolver link, observed 2026-07-13T05:02:03.008494Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.