Paper Citation Record · LEDGER

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability

As of 15 August 2026, this Paper Citation Record lists 23 of 23 outbound references and 0 inbound Pith citation observations for arXiv:2608.08904.

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

pith.paper-citation-record.v1
2608.08904 v1

Coverage vector

measured 23 of 23 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-14T04:32:44.985993Z

measured 23 of 23 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-15T06:32:42.880941+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

23 of 23 outbound references displayed

verified exact1
verified fuzzy4
unresolved18
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 6e75eabd-48d0-43e8-8b18-7f1c5fff0c8f · outbound

This paper cites Probing the 3D Awareness of Visual Foundation Models.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability Probing the 3D Awareness of Visual Foundation Models

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-14T04:32:44.909282Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation de8d0a9d-8e96-4500-8bce-be5c9d96a3a9 · outbound

This paper cites $\pi_0$: A Vision-Language-Action Flow Model for General Robot Control.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability $\pi_0$: A Vision-Language-Action Flow Model for General Robot Control

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-14T04:32:44.913405Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 28f13a92-ab94-490a-8e7b-a3f25fd027f5 · outbound

This paper cites an unresolved cited work.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability Unresolved cited work

Reference 3

Resolution: unresolved; raw_fallback, observed 2026-08-14T04:32:45.331055Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-15T06:32:42.880941+00:00.

Observation 51a1fff4-58ca-4464-a79d-b2e3f1c05e43 · outbound

This paper cites RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-14T04:32:44.920697Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a181b322-52df-4293-bcf8-dbf8ea7fd92a · outbound

This paper cites RT-1: Robotics Transformer for Real-World Control at Scale.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability RT-1: Robotics Transformer for Real-World Control at Scale

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-14T04:32:44.923942Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8208b93f-7d51-417b-a91c-9fafeb2c98ba · outbound

This paper cites Lawrence Erl- baum Associates, 2 edn.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability Lawrence Erl- baum Associates, 2 edn

Reference 6

Resolution: verified fuzzy; raw_fallback, observed 2026-08-14T04:32:45.322309Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-15T06:32:42.880941+00:00.

Observation 39400551-17af-4ba7-9d22-07501ad615d7 · outbound

This paper cites Knowledge Insulating Vision-Language-Action Models: Train Fast, Run Fast, Generalize Better.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability Knowledge Insulating Vision-Language-Action Models: Train Fast, Run Fast, Generalize Better

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-14T04:32:44.942639Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1a80fa48-5aba-4772-ae57-3b657eb25423 · outbound

This paper cites Transformer Circuits Thread (2021),https: //transformer-circuits.pub/2021/framework/index.html.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability Transformer Circuits Thread (2021),https: //transformer-circuits.pub/2021/framework/index.html

Reference 8

Resolution: verified fuzzy; raw_fallback, observed 2026-08-14T04:32:45.314255Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-15T06:32:42.880941+00:00.

Observation fcbce151-4299-4621-86de-814b021eae3d · outbound

This paper cites MolmoAct2: Action Reasoning Models for Real-world Deployment.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability MolmoAct2: Action Reasoning Models for Real-world Deployment

Reference 9

Resolution: unresolved; no resolver link, observed 2026-08-14T04:32:44.948419Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e5afcdb7-fb68-485c-9865-70b5fe39a1d7 · outbound

This paper cites Transformer Feed-Forward Layers Build Predictions by Promoting Concepts in the Vocabulary Space.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability Transformer Feed-Forward Layers Build Predictions by Promoting Concepts in the Vocabulary Space

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-14T04:32:44.951236Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4511cab6-af4e-47c6-8798-6de13e483da4 · outbound

This paper cites Transformer Feed-Forward Layers Are Key-Value Memories.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability Transformer Feed-Forward Layers Are Key-Value Memories

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-14T04:32:44.954186Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7cad8414-22dd-475d-a530-459462813768 · outbound

This paper cites an unresolved cited work.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability Unresolved cited work

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-14T04:32:44.957266Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1658f062-6ef7-46e5-a2ea-8dfe61dcab64 · outbound

This paper cites OpenVLA: An Open-Source Vision-Language-Action Model.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability OpenVLA: An Open-Source Vision-Language-Action Model

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-14T04:32:44.959556Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c8bd2a04-cf5a-4ff8-8820-69d53c7ac401 · outbound

This paper cites Depth Anything 3: Recovering the Visual Space from Any Views.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability Depth Anything 3: Recovering the Visual Space from Any Views

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-14T04:32:44.962378Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d04c7646-b0cd-4edd-8ddc-5cef1a831275 · outbound

This paper cites LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learning.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learning

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-14T04:32:44.964918Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6a713347-9697-4530-a832-2594a70333ec · outbound

This paper cites In: Advances in Neural Information Processing Systems (2024),https://arxiv.org/abs/2409.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability In: Advances in Neural Information Processing Systems (2024),https://arxiv.org/abs/2409

Reference 16

Resolution: verified fuzzy; raw_fallback, observed 2026-08-14T04:32:45.305805Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-15T06:32:42.880941+00:00.

Observation d75fd2e5-4cc8-49ba-93e9-e9b69d67b8f2 · outbound

This paper cites Locating and Editing Factual Associations in GPT.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability Locating and Editing Factual Associations in GPT

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-14T04:32:44.970047Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8fee45f9-c5ba-449b-a4d7-24ccbbfa5961 · outbound

This paper cites Octo: An Open-Source Generalist Robot Policy.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability Octo: An Open-Source Generalist Robot Policy

Reference 18

Resolution: unresolved; no resolver link, observed 2026-08-14T04:32:44.972658Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f350fbba-0654-4260-a30c-e636659901c2 · outbound

This paper cites FAST: Efficient Action Tokenization for Vision-Language-Action Models.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability FAST: Efficient Action Tokenization for Vision-Language-Action Models

Reference 19

Resolution: unresolved; no resolver link, observed 2026-08-14T04:32:44.975880Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 850fabfc-b14b-4658-ad72-f3e3ce984c60 · outbound

This paper cites Vision Transformers for Dense Prediction.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability Vision Transformers for Dense Prediction

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-14T04:32:44.978587Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9f1e0eb2-2afe-482c-8718-3036067ad2df · outbound

This paper cites SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-14T04:32:44.981103Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 71004ad8-82fd-4766-8a72-6710f1f3956d · outbound

This paper cites an unresolved cited work.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability Unresolved cited work

Reference 22

Resolution: verified exact; raw_fallback, observed 2026-08-14T04:32:45.061989Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-15T06:32:42.880941+00:00.

Observation 5b108007-62d8-479e-b38c-1b7b450cc849 · outbound

This paper cites The data are LIBERO frames sam- pled at stride 5 per rollout, resized to 256 pixels, with primary and wrist views.

From Recovery to Drop-off: How Action Post-training Reduces a VLM's Late-Layer Depth Decodability The data are LIBERO frames sam- pled at stride 5 per rollout, resized to 256 pixels, with primary and wrist views

Reference 23

Resolution: verified fuzzy; raw_fallback, observed 2026-08-14T04:32:45.297054Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-15T06:32:42.880941+00:00.

Pith citing papers

No inbound Pith citation observations are available.