Paper Citation Record · LEDGER

Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

As of 11 August 2026, this Paper Citation Record lists 14 of 14 outbound references and 100 inbound Pith citation observations for arXiv:2408.05147.

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

pith.paper-citation-record.v1
2408.05147 v2

Coverage vector

measured 14 of 14 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-07-11T11:50:26.030339Z

measured 114 of 114 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-10T06:31:04.303077+00:00

measured 100 of 103 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-08-10T22:18:23.605197Z

measured 1 of 1 external citation measurements

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

Source: pith, observed 2026-08-05T02:28:24.338817Z

Reference resolution

14 of 14 outbound references displayed

verified exact5
verified fuzzy4
unresolved3
parse uncertain0
malformed identifier0
metadata mismatch2

External citation measurements

1
pith, observed 2026-08-05T02:28:24.338817Z

Outbound references

Observation 82f206c9-eff3-4bc6-a355-f557798bbec4 · outbound

This paper cites Towards Automated Circuit Discovery for Mechanistic Interpretability.

Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2 Towards Automated Circuit Discovery for Mechanistic Interpretability

Reference 1

Resolution: metadata mismatch; arxiv_id, observed 2026-05-15T05:47:20.036422Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 0d8e817e-9593-441b-a882-0cbf6f7166de · outbound

This paper cites Gemini: A Family of Highly Capable Multimodal Models.

Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2 Gemini: A Family of Highly Capable Multimodal Models

Reference 2

Resolution: verified exact; local_arxiv, observed 2026-05-15T05:47:20.028841Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation d799f6df-df6a-4a56-8dec-6e051c7446b7 · outbound

This paper cites Universal Neurons in GPT2 Language Models.

Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2 Universal Neurons in GPT2 Language Models

Reference 3

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.049487Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation b64358ae-0f70-4f30-80f6-2c2230dbbc40 · outbound

This paper cites Adam: A Method for Stochastic Optimization.

Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2 Adam: A Method for Stochastic Optimization

Reference 4

Resolution: verified exact; local_arxiv, observed 2026-05-15T05:47:20.021747Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 5f156665-e25c-4aaf-a03c-85764d009081 · outbound

This paper cites Efficient Estimation of Word Representations in Vector Space.

Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2 Efficient Estimation of Word Representations in Vector Space

Reference 5

Resolution: metadata mismatch; local_arxiv, observed 2026-05-15T05:47:20.008082Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 30bd974b-b44d-4078-a44e-b2bedf7dca65 · outbound

This paper cites Distill , year =.

Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2 Distill , year =

Reference 6

Resolution: verified exact; doi, observed 2026-05-15T05:47:20.014372Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 73623492-eea4-4e80-a147-caa65323a6c5 · outbound

This paper cites Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small.

Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2 Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small

Reference 7

Resolution: verified exact; local_arxiv, observed 2026-05-15T05:47:20.043114Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation e80888af-3233-44f2-822f-56a50854ceed · outbound

This paper cites 17 Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2 Gemma 2 has Gated MLPs unlike GPT-2 Small) as well as SAEs.

Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2 17 Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2 Gemma 2 has Gated MLPs unlike GPT-2 Small) as well as SAEs

Reference 8

Resolution: verified fuzzy; raw_fallback, observed 2026-05-15T05:47:20.059685Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 51d25de6-c743-40f7-ad5a-126f0f7d109a · outbound

This paper cites an unresolved cited work.

Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2 Unresolved cited work

Reference 9

Resolution: unresolved; raw_fallback, observed 2026-05-15T05:47:20.064308Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation acf3606a-aa8d-4c3e-b38d-c3726ffb06ac · outbound

This paper cites an unresolved cited work.

Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2 Unresolved cited work

Reference 10

Resolution: unresolved; raw_fallback, observed 2026-05-15T05:47:20.069161Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation e07a8409-a5ea-4e63-a19f-9110ba1626cf · outbound

This paper cites Dunefsky et al.

Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2 Dunefsky et al

Reference 11

Resolution: verified fuzzy; raw_fallback, observed 2026-05-15T05:47:20.073612Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 96609cb6-c92f-47e4-b495-69aaed5bf4e8 · outbound

This paper cites Language model technical details We fold the pre-MLP RMS norm gain parameters (Zhang and Sennrich (2019), Section 3) into the MLP input matrices, as described in (Gurnee et al.

Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2 Language model technical details We fold the pre-MLP RMS norm gain parameters (Zhang and Sennrich (2019), Section 3) into the MLP input matrices, as described in (Gurnee et al

Reference 12

Resolution: verified fuzzy; raw_fallback, observed 2026-05-15T05:47:20.078971Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 1f5ba481-63c7-4ddd-90a7-d58f2165b6c4 · outbound

This paper cites an unresolved cited work.

Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2 Unresolved cited work

Reference 13

Resolution: unresolved; raw_fallback, observed 2026-05-15T05:47:20.083590Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 6b86dee1-9ab1-4833-8d2c-c69c134d28e9 · outbound

This paper cites google/gemma-2-2b.

Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2 google/gemma-2-2b

Reference 14

Resolution: verified fuzzy; raw_fallback, observed 2026-05-15T05:47:20.054641Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Pith citing papers

Observation 3962a406-2cc0-47e6-80aa-8ccc0d3ecc62 · inbound

METAGENE-1: Metagenomic Foundation Model for Pandemic Monitoring cites this paper.

METAGENE-1: Metagenomic Foundation Model for Pandemic Monitoring Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-10T22:18:23.605197Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation de2a5c91-fb68-4c4e-b887-448a37b15947 · inbound

Rethinking Evaluation of Sparse Autoencoders through the Representation of Polysemous Words cites this paper.

Rethinking Evaluation of Sparse Autoencoders through the Representation of Polysemous Words Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-10T21:28:15.942151Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d6d8fbe5-0347-4fd7-b6d7-14cc6fe30d10 · inbound

Sparse Autoencoders Trained on the Same Data Learn Different Features cites this paper.

Sparse Autoencoders Trained on the Same Data Learn Different Features Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-10T11:58:37.077293Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8bcccf64-82e2-4530-bbbd-df27963c3d67 · inbound

Low-Rank Adapting Models for Sparse Autoencoders cites this paper.

Low-Rank Adapting Models for Sparse Autoencoders Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-09T20:18:30.973761Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 295d5822-a12f-4c41-9406-d985907c68e6 · inbound

Sparse Autoencoder Insights on Voice Embeddings cites this paper.

Sparse Autoencoder Insights on Voice Embeddings Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 2024

Resolution: unresolved; no resolver link, observed 2026-08-09T20:10:32.789781Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3eaf50f9-ab4e-48a8-a132-b22b001828b1 · inbound

Analyze Feature Flow to Enhance Interpretation and Steering in Language Models cites this paper.

Analyze Feature Flow to Enhance Interpretation and Steering in Language Models Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-09T10:11:51.766856Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3815e31b-4fd3-4f2d-84e6-28efbfb6b15c · inbound

Sparse Autoencoders Do Not Find Canonical Units of Analysis cites this paper.

Sparse Autoencoders Do Not Find Canonical Units of Analysis Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-08T21:12:13.846998Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation fa921c34-69a1-4f13-bb2a-16033b96982f · inbound

Beyond External Monitors: Enhancing Transparency of Large Language Models for Easier Monitoring cites this paper.

Beyond External Monitors: Enhancing Transparency of Large Language Models for Easier Monitoring Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 47

Resolution: unresolved; no resolver link, observed 2026-08-08T21:06:55.960754Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4e812e43-1633-4be2-9ae8-9156c926c1da · inbound

LLM Pretraining with Continuous Concepts cites this paper.

LLM Pretraining with Continuous Concepts Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-08T04:54:39.330610Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 86dbdc20-721d-4972-aeb1-632c337a0e29 · inbound

Beyond Cross-Modal Alignment: Measuring and Leveraging Modality Gap in Vision-Language Models cites this paper.

Beyond Cross-Modal Alignment: Measuring and Leveraging Modality Gap in Vision-Language Models Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 18

Resolution: verified exact; local_arxiv, observed 2026-05-23T02:55:19.743502Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 13023214-b0bf-4312-be60-747afb101da2 · inbound

Towards eliciting latent knowledge from LLMs with mechanistic interpretability cites this paper.

Towards eliciting latent knowledge from LLMs with mechanistic interpretability Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 18

Resolution: unresolved; no resolver link, observed 2026-08-07T15:40:06.584693Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 201a46ff-68fa-4b96-95bd-71bb9cfebe6b · inbound

Denoising Concept Vectors with Sparse Autoencoders for Improved Language Model Steering cites this paper.

Denoising Concept Vectors with Sparse Autoencoders for Improved Language Model Steering Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-07T15:30:21.356728Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b6963240-da6c-4b85-9e44-ac92f5018386 · inbound

Feature Extraction and Steering for Enhanced Chain-of-Thought Reasoning in Language Models cites this paper.

Feature Extraction and Steering for Enhanced Chain-of-Thought Reasoning in Language Models Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-07T15:17:15.847691Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 921c1e0c-2bbc-40d4-ad40-2bfa05367e1a · inbound

Sparse Activation Editing for Reliable Instruction Following in Narratives cites this paper.

Sparse Activation Editing for Reliable Instruction Following in Narratives Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-07T15:04:54.767562Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bb2d5175-7072-4bf8-8535-571c78840d20 · inbound

Inference-Time Decomposition of Activations (ITDA): A Scalable Approach to Interpreting Large Language Models cites this paper.

Inference-Time Decomposition of Activations (ITDA): A Scalable Approach to Interpreting Large Language Models Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-07T14:44:42.856429Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 43462163-784f-44bf-9fc6-bb51c38f0bd3 · inbound

Factual Self-Awareness in Language Models: Representation, Robustness, and Scaling cites this paper.

Factual Self-Awareness in Language Models: Representation, Robustness, and Scaling Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 34

Resolution: unresolved; no resolver link, observed 2026-08-07T13:41:57.703508Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1a18718c-642b-4d93-bcfb-27f9fa7b5b9a · inbound

Understanding Refusal in Language Models with Sparse Autoencoders cites this paper.

Understanding Refusal in Language Models with Sparse Autoencoders Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 23

Resolution: unresolved; no resolver link, observed 2026-08-07T12:51:06.983803Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 804b3db4-8e99-435d-a6c0-0ffb280dc208 · inbound

Linear Representation Transferability Hypothesis: Leveraging Small Models to Steer Large Models cites this paper.

Linear Representation Transferability Hypothesis: Leveraging Small Models to Steer Large Models Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 18

Resolution: unresolved; no resolver link, observed 2026-08-07T12:07:10.502533Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 18c34387-d8b4-40a1-b02d-8d4dfa162506 · inbound

Incorporating Hierarchical Semantics in Sparse Autoencoder Architectures cites this paper.

Incorporating Hierarchical Semantics in Sparse Autoencoder Architectures Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-07T11:56:15.189378Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6190e1b7-d4ce-4800-9123-c51396bdc8d4 · inbound

BehaviorBox: Automated Discovery of Fine-Grained Performance Differences Between Language Models cites this paper.

BehaviorBox: Automated Discovery of Fine-Grained Performance Differences Between Language Models Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-07T11:32:31.922004Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e4aa4638-5bda-425b-8ba2-001026c5dad3 · inbound

HyperSteer: Activation Steering at Scale with Hypernetworks cites this paper.

HyperSteer: Activation Steering at Scale with Hypernetworks Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-07T11:11:18.493936Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9d704216-f085-43b4-ad3d-bf434267d1ac · inbound

Interpretation Meets Safety: A Survey on Interpretation Methods and Tools for Improving LLM Safety cites this paper.

Interpretation Meets Safety: A Survey on Interpretation Methods and Tools for Improving LLM Safety Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 28

Resolution: unresolved; no resolver link, observed 2026-08-07T10:24:26.412268Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 99ca24ad-20bb-4290-94dc-230070a0d686 · inbound

InverseScope: Scalable Activation Inversion for Interpreting Large Language Models cites this paper.

InverseScope: Scalable Activation Inversion for Interpreting Large Language Models Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-07T05:44:34.843022Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8c1c0a91-2136-4f85-8932-070813fb987f · inbound

Breaking the Block: Preserving Data Continuity to Train Superior SAEs for Instruct Models cites this paper.

Breaking the Block: Preserving Data Continuity to Train Superior SAEs for Instruct Models Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 2024

Resolution: unresolved; no resolver link, observed 2026-08-07T05:34:17.079488Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bc6efdc3-a151-4fdc-8221-4521bf9c9819 · inbound

How Visual Representations Map to Language Feature Space in Multimodal LLMs cites this paper.

How Visual Representations Map to Language Feature Space in Multimodal LLMs Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-07T01:06:07.108965Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation fc1d7192-3f62-4e02-bc7b-2d7da1688571 · inbound

AI shares emotion with humans across languages and cultures cites this paper.

AI shares emotion with humans across languages and cultures Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-07T04:46:01.200322Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2ea5ab97-a966-4569-b43e-1040cb7710ba · inbound

Large Reasoning Models are not thinking straight: on the unreliability of thinking trajectories cites this paper.

Large Reasoning Models are not thinking straight: on the unreliability of thinking trajectories Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-06T21:13:12.903707Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e207ae10-aca5-4ae7-bb5b-b35637abd0ca · inbound

Evaluating SAE interpretability without explanations cites this paper.

Evaluating SAE interpretability without explanations Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-06T18:24:51.490925Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cf8dedc7-26c3-46b1-97a6-22baaf200c9c · inbound

CytoSAE: Interpretable Cell Embeddings for Hematology cites this paper.

CytoSAE: Interpretable Cell Embeddings for Hematology Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-06T16:51:53.376696Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation eaa66d1c-5b33-4e63-bb8b-4d5c3b84a061 · inbound

Insights into a radiology-specialised multimodal large language model with sparse autoencoders cites this paper.

Insights into a radiology-specialised multimodal large language model with sparse autoencoders Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 29

Resolution: unresolved; no resolver link, observed 2026-08-06T16:40:42.550922Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation fb72a4dd-16ea-4280-b27f-ad96476289fb · inbound

Mammo-SAE: Interpreting Breast Cancer Concept Learning with Sparse Autoencoders cites this paper.

Mammo-SAE: Interpreting Breast Cancer Concept Learning with Sparse Autoencoders Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-06T15:42:41.152355Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ba8ef8f1-9b1d-48bd-96f6-4a8756cb9402 · inbound

On the transferability of Sparse Autoencoders for interpreting compressed models cites this paper.

On the transferability of Sparse Autoencoders for interpreting compressed models Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 25

Resolution: unresolved; no resolver link, observed 2026-08-06T15:24:45.755574Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f5b1c08f-59e9-4a98-a36a-f00849e83316 · inbound

Interpreting CFD Surrogates through Sparse Autoencoders cites this paper.

Interpreting CFD Surrogates through Sparse Autoencoders Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-06T15:23:46.349160Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5a19e2ca-fe04-4979-aa2a-b105f0bbeb6d · inbound

ExploreGS: Explorable 3D Scene Reconstruction with Virtual Camera Samplings and Diffusion Priors cites this paper.

ExploreGS: Explorable 3D Scene Reconstruction with Virtual Camera Samplings and Diffusion Priors Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 60

Resolution: unresolved; no resolver link, observed 2026-08-05T23:01:47.423261Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bf60ead4-74c7-401f-bd58-7027e581ca26 · inbound

Sealing The Backdoor: Unlearning Adversarial Text Triggers In Diffusion Models Using Knowledge Distillation cites this paper.

Sealing The Backdoor: Unlearning Adversarial Text Triggers In Diffusion Models Using Knowledge Distillation Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 49

Resolution: unresolved; no resolver link, observed 2026-08-05T18:47:45.256443Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2a59495b-c537-4f20-80a1-79adccabe81e · inbound

Safe-SAIL: Towards a Fine-grained Safety Landscape of Large Language Models via Sparse Autoencoder Interpretation Framework cites this paper.

Safe-SAIL: Towards a Fine-grained Safety Landscape of Large Language Models via Sparse Autoencoder Interpretation Framework Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 24

Resolution: verified exact; local_arxiv, observed 2026-05-18T18:16:43.696934Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 889197ce-1165-4e6c-aa70-4a2b7dce9c1c · inbound

Towards Atoms of Large Language Models cites this paper.

Towards Atoms of Large Language Models Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-04T15:20:32.174473Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 69c8943c-5e17-44d0-9191-3d878f769783 · inbound

ActivationReasoning: Logical Reasoning in Latent Activation Spaces cites this paper.

ActivationReasoning: Logical Reasoning in Latent Activation Spaces Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 10

Resolution: verified exact; local_arxiv, observed 2026-05-18T05:45:56.271876Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 1886ba7f-a6ee-420e-ab41-000b17af3f07 · inbound

Making Interpretable Discoveries from Unstructured Data: A High-Dimensional Multiple Hypothesis Testing Approach cites this paper.

Making Interpretable Discoveries from Unstructured Data: A High-Dimensional Multiple Hypothesis Testing Approach Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 53

Resolution: verified exact; local_arxiv, observed 2026-05-18T02:00:40.033494Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 1aa41a25-3ffb-47c1-b7eb-b8f014c7ae2c · inbound

Making Interpretable Discoveries from Unstructured Data: A High-Dimensional Multiple Hypothesis Testing Approach cites this paper.

Making Interpretable Discoveries from Unstructured Data: A High-Dimensional Multiple Hypothesis Testing Approach Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 53

Resolution: unresolved; no resolver link, observed 2026-08-04T00:23:30.786316Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b817608e-4c93-4d4a-8cf2-98ebe742fe88 · inbound

The SuperActivator Mechanism: Transformers Concentrate Reliable Concept Signals in the Tail cites this paper.

The SuperActivator Mechanism: Transformers Concentrate Reliable Concept Signals in the Tail Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 36

Resolution: unresolved; no resolver link, observed 2026-08-03T18:33:04.135713Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3e461bd4-07de-4407-92f4-26e81a05e337 · inbound

Less is Enough: Synthesizing Diverse Data in LLM Feature Space with Sparse Autoencoders cites this paper.

Less is Enough: Synthesizing Diverse Data in LLM Feature Space with Sparse Autoencoders Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 85

Resolution: unresolved; no resolver link, observed 2026-08-03T01:17:12.094097Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 92470775-d594-4e06-8be6-992f70fb6e7e · inbound

When Benchmarks Lie: Evaluating Malicious Prompt Classifiers Under True Distribution Shift cites this paper.

When Benchmarks Lie: Evaluating Malicious Prompt Classifiers Under True Distribution Shift Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-02T23:22:38.478153Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cd9fea74-b7bd-44d1-81b3-6e8e1e667939 · inbound

MetaSAEs: Joint Training with a Decomposability Penalty Produces More Atomic Sparse Autoencoder Latents cites this paper.

MetaSAEs: Joint Training with a Decomposability Penalty Produces More Atomic Sparse Autoencoder Latents Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 11

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 10472888-231d-40cb-9d3c-c1d31fea84c6 · inbound

Sparse Autoencoders as a Steering Basis for Phase Synchronization in Graph-Based CFD Surrogates cites this paper.

Sparse Autoencoders as a Steering Basis for Phase Synchronization in Graph-Based CFD Surrogates Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 5

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation de27df86-0ccb-4505-9f41-73b9fe159ddd · inbound

Geometric quantification for nonlinear deformation in knitted fabrics cites this paper.

Geometric quantification for nonlinear deformation in knitted fabrics Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 30

Resolution: unresolved; no resolver link, observed 2026-07-15T11:27:58.396234Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a8791cb6-b65a-4885-a773-50dc50709060 · inbound

SAGE: Signal-Amplified Guided Embeddings for LLM-based Vulnerability Detection cites this paper.

SAGE: Signal-Amplified Guided Embeddings for LLM-based Vulnerability Detection Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 30

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation e3ff0d7f-8a65-4407-94f9-448990d56d98 · inbound

Understanding the Mechanism of Altruism in Large Language Models cites this paper.

Understanding the Mechanism of Altruism in Large Language Models Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 235

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation f6fc003c-f2a8-4817-9e9e-1099580dc68b · inbound

Do Hallucination Neurons Generalize? Evidence from Cross-Domain Transfer in LLMs cites this paper.

Do Hallucination Neurons Generalize? Evidence from Cross-Domain Transfer in LLMs Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 27

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 28fc4626-c23d-4750-8897-a7d2e2992b12 · inbound

Are LLM Uncertainty and Correctness Encoded by the Same Features? A Functional Dissociation via Sparse Autoencoders cites this paper.

Are LLM Uncertainty and Correctness Encoded by the Same Features? A Functional Dissociation via Sparse Autoencoders Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 47

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 057bda72-707a-43ed-be8f-6d2563788569 · inbound

Linear-Readout Floors and Threshold Recovery in Computation in Superposition cites this paper.

Linear-Readout Floors and Threshold Recovery in Computation in Superposition Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 10

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 14882c30-e0ac-4a5e-a478-71e173662c0b · inbound

Bucketing the Good Apples: A Method for Diagnosing and Improving Causal Abstraction cites this paper.

Bucketing the Good Apples: A Method for Diagnosing and Improving Causal Abstraction Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 13

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation ce2f2ada-9eb4-4079-9e00-d1d668421f39 · inbound

Steering grids for sparse-autoencoder features: when a top-context label names an activation regime rather than a causal axis cites this paper.

Steering grids for sparse-autoencoder features: when a top-context label names an activation regime rather than a causal axis Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 11

Resolution: metadata mismatch; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 69a5a946-20d2-4b9e-8a88-4a26bd95095c · inbound

Steering grids for sparse-autoencoder features: when a top-context label names an activation regime rather than a causal axis cites this paper.

Steering grids for sparse-autoencoder features: when a top-context label names an activation regime rather than a causal axis Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-02T14:57:40.345836Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bd7ab742-ab0d-45ce-b492-bd318a053e61 · inbound

Manifold Steering Reveals the Shared Geometry of Neural Network Representation and Behavior cites this paper.

Manifold Steering Reveals the Shared Geometry of Neural Network Representation and Behavior Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 203

Resolution: metadata mismatch; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 8d607369-5204-4867-a46d-3ab3c4c32b6f · inbound

Structural Instability of Feature Composition cites this paper.

Structural Instability of Feature Composition Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 3

Resolution: metadata mismatch; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation ad417b8c-e4f4-4916-a3ea-f99323079d29 · inbound

Feature Starvation as Geometric Instability in Sparse Autoencoders cites this paper.

Feature Starvation as Geometric Instability in Sparse Autoencoders Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 24

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation f1a37b84-1022-4751-aa8c-b952dad92f25 · inbound

SLAM: Structural Linguistic Activation Marking for Language Models cites this paper.

SLAM: Structural Linguistic Activation Marking for Language Models Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 13

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation f9c5119d-3673-48a1-a7e4-9eb8279f6065 · inbound

Tree SAE: Learning Hierarchical Feature Structures in Sparse Autoencoders cites this paper.

Tree SAE: Learning Hierarchical Feature Structures in Sparse Autoencoders Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 14

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation fedbe4ef-d20b-495e-b390-27cbb1a2ef37 · inbound

Tree SAE: Learning Hierarchical Feature Structures in Sparse Autoencoders cites this paper.

Tree SAE: Learning Hierarchical Feature Structures in Sparse Autoencoders Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 14

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 35c467e4-a0df-48a3-bc69-053894d40c15 · inbound

Tool Calling is Linearly Readable and Steerable in Language Models cites this paper.

Tool Calling is Linearly Readable and Steerable in Language Models Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 57

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation d5a8f24c-e198-4779-ac15-81654ef3a347 · inbound

The Echo Amplifies the Knowledge: Somatic Marker Analogues in Language Models via Emotion Vector Re-Injection cites this paper.

The Echo Amplifies the Knowledge: Somatic Marker Analogues in Language Models via Emotion Vector Re-Injection Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 12

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation dd47730e-0799-409c-8b16-2313a264aec1 · inbound

DECO: Sparse Mixture-of-Experts with Dense-Comparable Performance on End-Side Devices cites this paper.

DECO: Sparse Mixture-of-Experts with Dense-Comparable Performance on End-Side Devices Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 148

Resolution: metadata mismatch; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 113eebf4-69b5-40e6-af5f-7d350345d89a · inbound

DECO: Sparse Mixture-of-Experts with Dense-Comparable Performance on End-Side Devices cites this paper.

DECO: Sparse Mixture-of-Experts with Dense-Comparable Performance on End-Side Devices Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 148

Resolution: metadata mismatch; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 3e755c05-0406-4e93-a38a-73809a61597b · inbound

DECO: Sparse Mixture-of-Experts with Dense-Comparable Performance on End-Side Devices cites this paper.

DECO: Sparse Mixture-of-Experts with Dense-Comparable Performance on End-Side Devices Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 148

Resolution: metadata mismatch; local_arxiv, observed 2026-05-21T07:59:50.287717Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 824fe888-d8da-424d-9d35-312c245884fb · inbound

WriteSAE: Sparse Autoencoders for Recurrent State cites this paper.

WriteSAE: Sparse Autoencoders for Recurrent State Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 38

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 1efab8f4-e020-4186-ba8f-8dbf08fba6cb · inbound

WriteSAE: Sparse Autoencoders for Recurrent State cites this paper.

WriteSAE: Sparse Autoencoders for Recurrent State Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 38

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 64b8e219-8511-4694-825b-006f8f2199f5 · inbound

WriteSAE: Sparse Autoencoders for Recurrent State cites this paper.

WriteSAE: Sparse Autoencoders for Recurrent State Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 26

Resolution: metadata mismatch; local_arxiv, observed 2026-05-21T07:49:50.240937Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 5fd8d261-d048-4529-8187-699728714e6a · inbound

Descriptive Collision in Sparse Autoencoder Auto-Interpretability: When One Explanation Describes Many Features cites this paper.

Descriptive Collision in Sparse Autoencoder Auto-Interpretability: When One Explanation Describes Many Features Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 17

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 6bed4314-7bfa-454a-8e1e-d9e159865c76 · inbound

Mechanistic Interpretability of EEG Foundation Models via Sparse Autoencoders cites this paper.

Mechanistic Interpretability of EEG Foundation Models via Sparse Autoencoders Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 12

Resolution: verified exact; arxiv_id, observed 2026-05-15T05:47:20.085339Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 17762393-a890-4d94-abd8-cafa6a0673bf · inbound

Mechanistic Interpretability of EEG Foundation Models via Sparse Autoencoders cites this paper.

Mechanistic Interpretability of EEG Foundation Models via Sparse Autoencoders Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 12

Resolution: verified exact; local_arxiv, observed 2026-05-19T16:52:40.031963Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 1d3bb9fd-e5e4-4ead-9bae-b0d12bb68004 · inbound

Mechanistic Interpretability of EEG Foundation Models via Sparse Autoencoders cites this paper.

Mechanistic Interpretability of EEG Foundation Models via Sparse Autoencoders Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 12

Resolution: verified exact; local_arxiv, observed 2026-05-25T06:06:42.796540Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 4258b428-fb97-4f27-885e-d0b7d90bfac9 · inbound

Beyond Linear Superposition: Discovering Climate Features in AI Weather Models with KAN-SAE cites this paper.

Beyond Linear Superposition: Discovering Climate Features in AI Weather Models with KAN-SAE Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 10

Resolution: verified exact; local_arxiv, observed 2026-05-20T14:48:23.202418Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 475c22f6-35ee-4019-a8a6-1ee2b67a9647 · inbound

Reading Task Failure Off the Activations: A Sparse-Feature Audit of GPT-2 Small on Indirect Object Identification cites this paper.

Reading Task Failure Off the Activations: A Sparse-Feature Audit of GPT-2 Small on Indirect Object Identification Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 6

Resolution: verified exact; local_arxiv, observed 2026-05-22T07:44:42.741349Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 6dddb91f-59c4-45a1-9b04-4b053ecc3487 · inbound

Polymorphism Is Rotation: Operational Mechanistic Interpretability from a Two-Layer Transformer to Pythia-70m cites this paper.

Polymorphism Is Rotation: Operational Mechanistic Interpretability from a Two-Layer Transformer to Pythia-70m Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 9

Resolution: verified exact; local_arxiv, observed 2026-06-30T14:14:45.503377Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 48ab9548-bb9a-486d-9198-5620af97d852 · inbound

Sign-Aware Gated Sparse Autoencoders: Modeling Anticorrelated Features with Bi-Jump-ReLU Activations cites this paper.

Sign-Aware Gated Sparse Autoencoders: Modeling Anticorrelated Features with Bi-Jump-ReLU Activations Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 18

Resolution: verified exact; local_arxiv, observed 2026-06-29T14:23:30.741279Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 09ed6fe4-5176-4d34-99fb-2c663ef756b6 · inbound

Sign-Aware Gated Sparse Autoencoders: Modeling Anticorrelated Features with Bi-Jump-ReLU Activations cites this paper.

Sign-Aware Gated Sparse Autoencoders: Modeling Anticorrelated Features with Bi-Jump-ReLU Activations Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-04T05:02:51.319463Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7ec33387-212e-4c05-a071-246ab361e2e8 · inbound

Mechanistically Interpreting the Role of Sample Difficulty in RLVR for LLMs cites this paper.

Mechanistically Interpreting the Role of Sample Difficulty in RLVR for LLMs Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 22

Resolution: verified exact; local_arxiv, observed 2026-06-29T12:13:26.551175Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 9dc3e8fd-bee9-4afb-8721-079dcf2b76e3 · inbound

Interpretability-Guided Layer Selection over Subspace Projection: SAEs as Stethoscopes, Not Scalpels, for Raw Task Vector Model Editing cites this paper.

Interpretability-Guided Layer Selection over Subspace Projection: SAEs as Stethoscopes, Not Scalpels, for Raw Task Vector Model Editing Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 5

Resolution: verified exact; local_arxiv, observed 2026-06-29T14:03:29.441007Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 22ae2eea-52f2-4df6-b4fb-7a60ec7e476d · inbound

Feature Geometry of LoRA Adapters: A Sparse Autoencoder Analysis of Representational Divergence in Fine-Tuned Language Models cites this paper.

Feature Geometry of LoRA Adapters: A Sparse Autoencoder Analysis of Representational Divergence in Fine-Tuned Language Models Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 9

Resolution: verified exact; local_arxiv, observed 2026-06-29T13:53:28.773158Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 95aa11b0-e6c2-46ba-bd2f-8de029b083d8 · inbound

When and How Long? The Readout-Mediator Angle in Temporal Reasoning cites this paper.

When and How Long? The Readout-Mediator Angle in Temporal Reasoning Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 5

Resolution: verified exact; local_arxiv, observed 2026-06-29T13:33:27.895509Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation e07d41e3-2ad1-441a-ad33-48c2ac739dfe · inbound

BioRefusalAudit: Auditing Biosecurity Refusal Depth Using General and Domain-Fine-Tuned Sparse Autoencoders cites this paper.

BioRefusalAudit: Auditing Biosecurity Refusal Depth Using General and Domain-Fine-Tuned Sparse Autoencoders Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 4

Resolution: verified exact; local_arxiv, observed 2026-06-29T07:33:13.791646Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 9be6e381-1c13-4cf3-ab53-f870dd84b6a0 · inbound

Auditing Asset-Specific Preferences in Financial Large Language Models: Evidence from Bitcoin Representations and Portfolio Allocation cites this paper.

Auditing Asset-Specific Preferences in Financial Large Language Models: Evidence from Bitcoin Representations and Portfolio Allocation Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 2

Resolution: metadata mismatch; local_arxiv, observed 2026-06-28T11:52:04.964118Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation edc86375-2c7d-40c5-a282-9247a60f965d · inbound

Auditing Asset-Specific Preferences in Financial Large Language Models: Evidence from Bitcoin Representations and Portfolio Allocation cites this paper.

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-02T12:37:23.244657Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 72ab9113-4f24-43a5-819b-c2a6af1dbf8e · inbound

Characterize Then Distill: Mechanistic Reasoning in Large Output Spaces cites this paper.

Characterize Then Distill: Mechanistic Reasoning in Large Output Spaces Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 49

Resolution: verified exact; local_arxiv, observed 2026-06-27T22:31:21.402678Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation e67ec7d6-ae13-4768-a2fa-679ecf3d1018 · inbound

Interpreting and Steering a Text-to-Speech Language Model with Sparse Autoencoders cites this paper.

Interpreting and Steering a Text-to-Speech Language Model with Sparse Autoencoders Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 5

Resolution: metadata mismatch; local_arxiv, observed 2026-07-03T00:27:30.047693Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 7adaa47e-88b2-4e3e-82db-10fcffb79d49 · inbound

When the Chain of Thought Knows Better: Failure Modes in Multi-Turn Reasoning Models cites this paper.

When the Chain of Thought Knows Better: Failure Modes in Multi-Turn Reasoning Models Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 22

Resolution: metadata mismatch; local_arxiv, observed 2026-07-03T05:07:38.578013Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 58d4f9a5-2c95-452d-835e-1cf816787e8b · inbound

SAERec: Constructing Fine-grained Interpretable Intents Priors via Sparse Autoencoders for Recommendation cites this paper.

SAERec: Constructing Fine-grained Interpretable Intents Priors via Sparse Autoencoders for Recommendation Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 26

Resolution: verified exact; local_arxiv, observed 2026-07-04T02:39:25.077542Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation c633fb62-4121-4d0f-8a9e-72d7ec2a7f1e · inbound

HydraHead: From Head-Level Functional Heterogeneity to Specialized Attention Hybridization cites this paper.

HydraHead: From Head-Level Functional Heterogeneity to Specialized Attention Hybridization Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 20

Resolution: verified exact; local_arxiv, observed 2026-07-04T03:59:33.694608Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation d0e1b7b7-b8c7-4693-beaf-360a37dad8d6 · inbound

Evaluation Awareness Is Not One Capability: Evidence from Open Language Models cites this paper.

Evaluation Awareness Is Not One Capability: Evidence from Open Language Models Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 7

Resolution: metadata mismatch; local_arxiv, observed 2026-07-04T10:49:46.879693Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation a2989815-f997-403d-95fa-5eb1d4ce4c62 · inbound

Perfect Detection, Failed Control: The Geometry of Knowing vs. Steering in Language Models cites this paper.

Perfect Detection, Failed Control: The Geometry of Knowing vs. Steering in Language Models Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 5

Resolution: verified exact; local_arxiv, observed 2026-07-04T16:49:57.474500Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 896a8bf4-e5b9-4d4e-8e56-2bcd152bfbb6 · inbound

At the Edge of Understanding: Sparse Autoencoders Trace The Limits of Transformer Generalization cites this paper.

At the Edge of Understanding: Sparse Autoencoders Trace The Limits of Transformer Generalization Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 38

Resolution: verified exact; local_arxiv, observed 2026-06-26T01:28:50.555158Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 21eb3188-9f3c-4fa5-bae2-319f23668a37 · inbound

Where Do Models Find Happiness? Emotion Vectors in Open-Source LLMs cites this paper.

Where Do Models Find Happiness? Emotion Vectors in Open-Source LLMs Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 5

Resolution: metadata mismatch; local_arxiv, observed 2026-07-04T13:39:50.852419Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation cc3f4371-734a-4c75-981e-c2a85a452f47 · inbound

Forecasting With LLMs: Improved Generalization Through Feature Steering cites this paper.

Forecasting With LLMs: Improved Generalization Through Feature Steering Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 31

Resolution: verified exact; local_arxiv, observed 2026-07-04T13:59:52.514820Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation adf54146-71ed-4816-babe-44fca0692e00 · inbound

PRA-RAG: Provably Robust Aggregation in Retrieval-Augmented Generation against Retrieval Corruption cites this paper.

PRA-RAG: Provably Robust Aggregation in Retrieval-Augmented Generation against Retrieval Corruption Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 55

Resolution: metadata mismatch; local_arxiv, observed 2026-07-02T23:47:27.203806Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 33070664-df44-465c-8bc2-6568bd123a7b · inbound

Brand-as-Memory: Vision-Language Models Encode Causal, Mechanistically Localizable Credibility Priors for News Sources cites this paper.

Brand-as-Memory: Vision-Language Models Encode Causal, Mechanistically Localizable Credibility Priors for News Sources Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 10

Resolution: unresolved; no resolver link, observed 2026-07-12T03:00:02.282468Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7ea73d2a-9bc0-4859-993c-1d5d69319310 · inbound

Mechanistic Interpretability for Neural Networks: Circuits, Sparse Features and Symbolic Reasoning cites this paper.

Mechanistic Interpretability for Neural Networks: Circuits, Sparse Features and Symbolic Reasoning Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 36

Resolution: verified exact; local_arxiv, observed 2026-07-09T15:06:18.005735Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 81b2608a-47c8-4ec4-9b14-1a2c108bec51 · inbound

When Are Sparse Feature Interventions Actually Localized? Matched Evaluation for SAE-Based Safety Control cites this paper.

When Are Sparse Feature Interventions Actually Localized? Matched Evaluation for SAE-Based Safety Control Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 10

Resolution: unresolved; no resolver link, observed 2026-07-14T13:23:22.708604Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f4674fb3-55c8-465d-a1cd-102d210b04f1 · inbound

Retrieval is Enough: Training-Free Interpretability with a Tool-Using Agent cites this paper.

Retrieval is Enough: Training-Free Interpretability with a Tool-Using Agent Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-01T20:58:52.501832Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2a97c8cf-8c52-48a0-86e4-747c97ba3974 · inbound

Attention Without Grounding: Causal Evaluation of Visual Explanations in Medical VLMs cites this paper.

Attention Without Grounding: Causal Evaluation of Visual Explanations in Medical VLMs Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-01T15:04:52.056618Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.