Paper Citation Record · LEDGER

Latent Adversarial Training Improves the Representation of Refusal

As of 20 August 2026, this Paper Citation Record lists 18 of 18 outbound references and 0 inbound Pith citation observations for arXiv:2504.18872.

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

pith.paper-citation-record.v1
2504.18872 v1

Coverage vector

measured 18 of 18 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-16T10:11:37.884734Z

measured 18 of 18 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-20T06:33:59.587034+00:00

measured 0 of 0 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links

measured 0 of 1 external citation measurements

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

Source: cited_works

Reference resolution

18 of 18 outbound references displayed

verified exact0
verified fuzzy3
unresolved15
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation e8e19bdc-4eaf-4e62-94f0-8525ee4d0b8e · outbound

This paper cites Refusal in Language Models Is Mediated by a Single Direction.

Latent Adversarial Training Improves the Representation of Refusal Refusal in Language Models Is Mediated by a Single Direction

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-16T10:11:37.822570Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1f07c3d3-e847-4657-82db-9a88eb1c6366 · outbound

This paper cites latent\_adversarial\_training, 2024.

Latent Adversarial Training Improves the Representation of Refusal latent\_adversarial\_training, 2024

Reference 2

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T10:11:38.085267Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-20T06:33:59.587034+00:00.

Observation 99f24ba7-0d12-4217-ba35-2dca7db00136 · outbound

This paper cites Defending Against Unforeseen Failure Modes with Latent Adversarial Training.

Latent Adversarial Training Improves the Representation of Refusal Defending Against Unforeseen Failure Modes with Latent Adversarial Training

Reference 3

Resolution: unresolved; no resolver link, observed 2026-08-16T10:11:37.831020Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9cfd8403-661a-4daa-9039-b58b17411291 · outbound

This paper cites Mechanistically analyzing the effects of fine-tuning on procedurally defined tasks.

Latent Adversarial Training Improves the Representation of Refusal Mechanistically analyzing the effects of fine-tuning on procedurally defined tasks

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-16T10:11:37.835041Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d50944d6-633b-422e-a924-cfd51cd52358 · outbound

This paper cites LoRA Fine-tuning Efficiently Undoes Safety Training in Llama 2-Chat 70B.

Latent Adversarial Training Improves the Representation of Refusal LoRA Fine-tuning Efficiently Undoes Safety Training in Llama 2-Chat 70B

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-16T10:11:37.839207Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 81d0c4c4-fdd1-4eec-be1b-9523cb1459d7 · outbound

This paper cites Llama 2 7B Chat , 2023.

Latent Adversarial Training Improves the Representation of Refusal Llama 2 7B Chat , 2023

Reference 6

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T10:11:38.072677Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-20T06:33:59.587034+00:00.

Observation 09070545-108f-4b8a-81a9-b6b6dd6e5941 · outbound

This paper cites The Llama 3 Herd of Models.

Latent Adversarial Training Improves the Representation of Refusal The Llama 3 Herd of Models

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-16T10:11:37.846329Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6902d673-ae34-418d-ad02-370b873474d0 · outbound

This paper cites GPT-4 Technical Report.

Latent Adversarial Training Improves the Representation of Refusal GPT-4 Technical Report

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-16T10:11:37.849545Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 660e8bfb-2a4d-465c-87f0-5d2169bb1d0d · outbound

This paper cites Steering Llama 2 via Contrastive Activation Addition.

Latent Adversarial Training Improves the Representation of Refusal Steering Llama 2 via Contrastive Activation Addition

Reference 9

Resolution: unresolved; no resolver link, observed 2026-08-16T10:11:37.853271Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1af2426f-642e-4353-96d0-a7c20bfc3c0f · outbound

This paper cites Latent Adversarial Training Improves Robustness to Persistent Harmful Behaviors in LLMs.

Latent Adversarial Training Improves the Representation of Refusal Latent Adversarial Training Improves Robustness to Persistent Harmful Behaviors in LLMs

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-16T10:11:37.856549Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 43855377-9f47-48f7-8d99-cf4afd16afe2 · outbound

This paper cites Stanford Alpaca: An Instruction-following LLaMA model , 2023.

Latent Adversarial Training Improves the Representation of Refusal Stanford Alpaca: An Instruction-following LLaMA model , 2023

Reference 11

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T10:11:38.062124Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-20T06:33:59.587034+00:00.

Observation f7721851-abe9-4320-a3f0-961e43d5e154 · outbound

This paper cites Llama 2: Open Foundation and Fine-Tuned Chat Models.

Latent Adversarial Training Improves the Representation of Refusal Llama 2: Open Foundation and Fine-Tuned Chat Models

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-16T10:11:37.862434Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 61ff8dc4-fde5-4294-86d1-12f67372e810 · outbound

This paper cites Shadow Alignment: The Ease of Subverting Safely-Aligned Language Models.

Latent Adversarial Training Improves the Representation of Refusal Shadow Alignment: The Ease of Subverting Safely-Aligned Language Models

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-16T10:11:37.865408Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 41fd8d41-096d-4924-9ef0-80841f02c37b · outbound

This paper cites Universal and Transferable Adversarial Attacks on Aligned Language Models.

Latent Adversarial Training Improves the Representation of Refusal Universal and Transferable Adversarial Attacks on Aligned Language Models

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-16T10:11:37.868113Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 00d2d191-ac13-4e64-8277-4ae63aa08ec0 · outbound

This paper cites write newline.

Latent Adversarial Training Improves the Representation of Refusal write newline

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-16T10:11:37.872694Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2cb46f6b-b291-49cd-99c9-16cbbae715ca · outbound

This paper cites @esa (Ref.

Latent Adversarial Training Improves the Representation of Refusal @esa (Ref

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-16T10:11:37.877140Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 13d0334d-6335-4d5b-9187-743c72c641c1 · outbound

This paper cites an unresolved cited work.

Latent Adversarial Training Improves the Representation of Refusal Unresolved cited work

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-16T10:11:37.880913Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation dcde0731-1e3b-476f-8bdb-d49d5ea73e4f · outbound

This paper cites an unresolved cited work.

Latent Adversarial Training Improves the Representation of Refusal Unresolved cited work

Reference 18

Resolution: unresolved; no resolver link, observed 2026-08-16T10:11:37.884734Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

No inbound Pith citation observations are available.