Typed states for the displayed outbound observations.
Source: paper_references, paper_reference_links, observed 2026-05-10T08:31:53.114772Z
Paper Citation Record · LEDGER
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases
As of 4 August 2026, this Paper Citation Record lists 100 of 117 outbound references and 1 inbound Pith citation observation for arXiv:2604.16286.
A citation records a reference. It does not transfer a finding from one paper to another.
Coverage vector
One-hop event checks from named stored sources.
Source: scholarly_work_events, retraction_status_cache, observed 2026-08-03T06:30:56.289259+00:00
Pith citing papers itemized under the disclosed page cap.
Source: paper_references, paper_reference_links, observed 2026-08-01T12:49:57.400820Z
A source-named dated measurement, never combined with another source.
Source: cited_works
Reference resolution
100 of 117 outbound references displayed
- verified exact0
- verified fuzzy82
- unresolved18
- parse uncertain0
- malformed identifier0
- metadata mismatch0
External citation measurements
No source-named external measurement is stored.
Outbound references
Observation 19a71e41-587a-4fcc-ac10-de8d404e1291 · outbound
This paper cites Jimenez, John Yang, Alexander Wettig, Shunyu Yao, Kexin Pei, Ofir Press, and Karthik Narasimhan.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Jimenez, John Yang, Alexander Wettig, Shunyu Yao, Kexin Pei, Ofir Press, and Karthik Narasimhan
Reference 1
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.170855Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 9546270b-95b0-4968-99b5-a26f0ca2fe4f · outbound
This paper cites RE-bench: Evaluating frontier AI R&D capabilities of language model agents against human experts.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases RE-bench: Evaluating frontier AI R&D capabilities of language model agents against human experts
Reference 2
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.992059Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 250e1c7f-0b4f-4cf6-95f1-b07a03d0b67a · outbound
This paper cites legitimate AI researcher.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases legitimate AI researcher
Reference 3
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.079244Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 609adfa3-a98c-4c78-af36-9fc85849dd46 · outbound
This paper cites The adolescence of technology.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases The adolescence of technology
Reference 4
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.051102Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation a7426852-beb8-4b09-aa42-6edf1ceaac03 · outbound
This paper cites Bowman, and Evan Hubinger.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Bowman, and Evan Hubinger
Reference 5
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.110431Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 1cdaae8b-f94c-4ba4-b870-ac76305e5aba · outbound
This paper cites Emergent misalignment: Narrow finetuning can produce broadly misaligned LLMs.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Emergent misalignment: Narrow finetuning can produce broadly misaligned LLMs
Reference 6
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.271230Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 3ac2e9ff-bb0e-48a9-bb85-de220662d11e · outbound
This paper cites Stress testing deliberative alignment for anti-scheming training.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Stress testing deliberative alignment for anti-scheming training
Reference 7
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.044333Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation c6fa2eef-eb05-454a-9e7e-8ef2729715c2 · outbound
This paper cites Bowman, Misha Wagner, Fabien Roger, and Holden Karnofsky.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Bowman, Misha Wagner, Fabien Roger, and Holden Karnofsky
Reference 8
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.233981Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 851cc0c3-4c74-445d-a117-5a7a69ee5e3c · outbound
This paper cites Sabotage risk report: Claude Opus 4.6.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Sabotage risk report: Claude Opus 4.6
Reference 9
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.252781Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation df986c1e-01f9-4d1d-b067-504c3abd10a3 · outbound
This paper cites Alignment risk update: Claude Mythos Preview.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Alignment risk update: Claude Mythos Preview
Reference 10
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.028922Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 61f3b045-2038-4911-aa63-bfa8ff8d4167 · outbound
This paper cites CTRL-ALT-DECEIT: Sabotage evaluations for automated AI R&D.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases CTRL-ALT-DECEIT: Sabotage evaluations for automated AI R&D
Reference 11
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.936444Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 70f70870-772a-4922-be76-5514b5f564d6 · outbound
This paper cites Bowman, and David Duvenaud.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Bowman, and David Duvenaud
Reference 12
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.153863Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation c4aef78a-f00d-4b59-bca6-cd407f0131db · outbound
This paper cites Subliminal learning: Language models transmit behavioral traits via hidden signals in data.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Subliminal learning: Language models transmit behavioral traits via hidden signals in data
Reference 13
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.121546Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 5da021e6-050e-44a4-b80e-4fc03b2083e6 · outbound
This paper cites CoT red-handed: Stress testing chain-of-thought monitoring.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases CoT red-handed: Stress testing chain-of-thought monitoring
Reference 14
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.865417Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 530e3ae8-1613-4772-bd82-1cb86706b6cb · outbound
This paper cites Disentangling feature and lazy training in deep neural networks.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Disentangling feature and lazy training in deep neural networks
Reference 15
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.965334Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation fbb7052b-4e59-4d32-8c3d-497e44132958 · outbound
This paper cites Steering evaluation-aware language models to act like they are deployed.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Steering evaluation-aware language models to act like they are deployed
Reference 16
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.003796Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 073f4334-c7c5-4767-9110-43576e4d5fa4 · outbound
This paper cites Lessons from studying two-hop latent reasoning.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Lessons from studying two-hop latent reasoning
Reference 17
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.274882Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 741105fd-1611-4683-a45c-3f8c927e4a16 · outbound
This paper cites Copy suppression: Comprehensively understanding an attention head.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Copy suppression: Comprehensively understanding an attention head
Reference 18
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.018013Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation eff825d5-ab34-4918-957e-a6a2aebf687e · outbound
This paper cites Hidden in plain text: Emergence & mitigation of steganographic collusion in LLMs.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Hidden in plain text: Emergence & mitigation of steganographic collusion in LLMs
Reference 19
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.089937Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation b4cba413-d23e-4481-b869-ce7fa7538528 · outbound
This paper cites Multi-turn jailbreaks are simpler than they seem.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Multi-turn jailbreaks are simpler than they seem
Reference 20
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.241271Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 9a6c4e4f-65eb-46ee-8b76-bbf13805b0d0 · outbound
This paper cites AI control: Improving safety despite intentional subversion.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases AI control: Improving safety despite intentional subversion
Reference 21
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.203553Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation c5bbdc21-af17-421f-ad28-426a13201d5b · outbound
This paper cites Ctrl-z: Controlling AI agents via resampling.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Ctrl-z: Controlling AI agents via resampling
Reference 22
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.911054Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation cf2397ff-30d1-49c0-84f6-fee957ba15ac · outbound
This paper cites Guan, Aleksander Madry, Wojciech Zaremba, Jakub Pachocki, and David Farhi.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Guan, Aleksander Madry, Wojciech Zaremba, Jakub Pachocki, and David Farhi
Reference 23
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.157825Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation ece28bf3-d01f-4d8f-9947-f8e95b2c01da · outbound
This paper cites Training fails to elicit subtle reasoning in current language models.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Training fails to elicit subtle reasoning in current language models
Reference 24
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.149787Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 57b1ced0-1c33-4c3a-b266-f8cf301dc5d4 · outbound
This paper cites Factor(T,U): Factored cognition strengthens monitoring of untrusted AI.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Factor(T,U): Factored cognition strengthens monitoring of untrusted AI
Reference 25
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.174783Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation c21eb7a6-fc13-4f07-9e35-88b112cfd7d9 · outbound
This paper cites Basic legibility protocols improve trusted monitoring.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Basic legibility protocols improve trusted monitoring
Reference 26
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.057860Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 2fc9ebe3-54ad-4dd2-8630-eaebf179a9bf · outbound
This paper cites BashArena: A control setting for highly privileged AI agents.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases BashArena: A control setting for highly privileged AI agents
Reference 27
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.136090Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation abbc1177-2322-4e4d-b525-bf95a8a268ee · outbound
This paper cites SHADE-arena: Evaluating sabotage and monitoring in LLM agents.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases SHADE-arena: Evaluating sabotage and monitoring in LLM agents
Reference 28
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.881897Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation bc882448-7527-4698-9efc-0aff6019229d · outbound
This paper cites Adaptive deployment of untrusted LLMs reduces distributed threats.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Adaptive deployment of untrusted LLMs reduces distributed threats
Reference 29
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.893643Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 922a6c75-8be2-4e43-9375-cf98bdf81eff · outbound
This paper cites Brown, and Francis Rhys Ward.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Brown, and Francis Rhys Ward
Reference 30
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.210586Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation a315946d-d243-4c95-bc6a-85bc44632058 · outbound
This paper cites Sandbagging in agentic ML research.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Sandbagging in agentic ML research
Reference 31
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.900862Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation c4b3deac-0e60-4e3f-a947-a36cbbdcb9c9 · outbound
This paper cites LLM critics help catch LLM bugs.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases LLM critics help catch LLM bugs
Reference 32
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.869145Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation c02770aa-2ed6-4a8c-bd78-0f4db43a8fc3 · outbound
This paper cites A practical approach to verifying code at scale.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases A practical approach to verifying code at scale
Reference 33
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.105195Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 74e97300-6054-43bd-a978-ea0bddc71e24 · outbound
This paper cites IRIS: LLM-assisted static analysis for detecting security vulnerabilities.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases IRIS: LLM-assisted static analysis for detecting security vulnerabilities
Reference 34
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.164721Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation c99a00c8-d4c9-4d7c-a8f2-cbae14bc0a30 · outbound
This paper cites JITVul: A benchmark for evaluating LLM-based vulnerability detection in real-world code.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases JITVul: A benchmark for evaluating LLM-based vulnerability detection in real-world code
Reference 35
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.932832Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 7a22f65c-55f6-4879-bee9-ce96d409f2f9 · outbound
This paper cites From large to mammoth: A systematic evaluation of LLM architectures and quantization for vulnerability detection.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases From large to mammoth: A systematic evaluation of LLM architectures and quantization for vulnerability detection
Reference 36
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.007309Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 1c68cfd4-44c3-4d80-8f4b-d6540dc76ce0 · outbound
This paper cites Mythos preview cybersecurity report.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Mythos preview cybersecurity report
Reference 37
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.064871Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 2d54058e-4653-40e0-8113-07e0aa6dbca9 · outbound
This paper cites Leakage and the reproducibility crisis in ML-based science.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Leakage and the reproducibility crisis in ML-based science
Reference 38
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.093497Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation be9ac55d-fb1d-4c26-b602-8e9e5f83d390 · outbound
This paper cites Are we really making much progress? A worrying analysis of recent neural recommendation approaches.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Are we really making much progress? A worrying analysis of recent neural recommendation approaches
Reference 39
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.982224Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 361026f0-c738-49fa-9f22-5fb0b85fa144 · outbound
This paper cites Are GANs created equal? A large-scale study.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Are GANs created equal? A large-scale study
Reference 40
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.025455Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation e1fe6d93-79ff-48ba-8533-dced9200d656 · outbound
This paper cites On the state of the art of evaluation in neural language models.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases On the state of the art of evaluation in neural language models
Reference 41
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.897145Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 4252fa6d-23da-4ab1-ab58-e2cc1eb875d6 · outbound
This paper cites A metric learning reality check.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases A metric learning reality check
Reference 42
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.889813Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 1dd0debb-6d4f-4ee3-81b9-35851b60a188 · outbound
This paper cites Pitfalls of graph neural network evaluation.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Pitfalls of graph neural network evaluation
Reference 43
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.075734Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 32d29912-5a49-4206-8cfd-1374426c5be4 · outbound
This paper cites What is the state of neural network pruning? InProceedings of Machine Learning and Systems (MLSys).
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases What is the state of neural network pruning? InProceedings of Machine Learning and Systems (MLSys)
Reference 44
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.040908Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation c97c72a0-75ae-4901-bdfe-399485c59d0a · outbound
This paper cites Towards evaluating the robustness of neural networks.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Towards evaluating the robustness of neural networks
Reference 45
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.264829Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 7e6903c2-86d8-4993-94e4-b342253ef19d · outbound
This paper cites Obfuscated gradients give a false sense of security: Circumventing defenses to adversarial examples.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Obfuscated gradients give a false sense of security: Circumventing defenses to adversarial examples
Reference 46
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.975604Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation d573282a-6354-495c-8583-95afa23e3ed3 · outbound
This paper cites On adaptive attacks to adversarial example defenses.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases On adaptive attacks to adversarial example defenses
Reference 47
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.032926Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation cf21cb4e-a0f7-4582-89d1-5fbcf43cf515 · outbound
This paper cites an unresolved cited work.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Unresolved cited work
Reference 48
- Resolution
- unresolved
- raw_fallback, observed 2026-05-21T00:14:17.904124Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation d65179db-be13-4d8c-993b-9d1c73d08cd3 · outbound
This paper cites A step toward quantifying independently reproducible machine learning research.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases A step toward quantifying independently reproducible machine learning research
Reference 49
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.244808Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation d57e395e-cbf4-41b8-a07c-a0d02a8199ee · outbound
This paper cites Emergent world models and latent variable estimation in chess-playing language models.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Emergent world models and latent variable estimation in chess-playing language models
Reference 50
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.988578Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 58063378-f7c5-40a2-8f42-8931e7da98f9 · outbound
This paper cites Can language models learn to skip steps? 2025.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Can language models learn to skip steps? 2025
Reference 51
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.185508Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation a023374b-a327-458e-af06-df594d51ccac · outbound
This paper cites Reasoning models sometimes output illegible chains of thought.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Reasoning models sometimes output illegible chains of thought
Reference 52
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.958139Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 9a8953ca-805f-4264-bd48-68885235d899 · outbound
This paper cites subliminal learning reliably transfers preferences.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases subliminal learning reliably transfers preferences
Reference 53
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.925058Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 0d8ac9a9-1a0b-42a8-9bf1-4fca793a7595 · outbound
This paper cites monitor notice.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases monitor notice
Reference 54
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.082602Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation dc90355b-f53e-4877-b271-f25ba43032f3 · outbound
This paper cites lazy training.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases lazy training
Reference 55
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.132350Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 32d8873a-9fb2-4a85-bef9-b954b12bbda2 · outbound
This paper cites steering partially recovers deployment behavior.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases steering partially recovers deployment behavior
Reference 56
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.222270Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation d41e52e4-9502-4516-95ab-4cce8e494bfd · outbound
This paper cites demonstrated.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases demonstrated
Reference 57
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.196438Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 61009a6f-a30c-4e22-9a80-48220ef61cea · outbound
This paper cites That is, it suppresses the tendency for LLMs to copy tokens that have already appeared in-context.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases That is, it suppresses the tendency for LLMs to copy tokens that have already appeared in-context
Reference 58
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.261011Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 1e351684-f079-4a8c-a2ca-9bcd7075d025 · outbound
This paper cites Specifically, theylookat WUWL10H7 QKMLP0(WE).
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Specifically, theylookat WUWL10H7 QKMLP0(WE)
Reference 59
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.142868Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation b2f2ae17-6de0-4898-97fd-b632f24d65cb · outbound
This paper cites an unresolved cited work.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Unresolved cited work
Reference 60
- Resolution
- unresolved
- raw_fallback, observed 2026-05-21T00:14:18.226551Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 40ff1000-c4fc-4ada-8577-55e26a5ee7d6 · outbound
This paper cites an unresolved cited work.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Unresolved cited work
Reference 61
- Resolution
- unresolved
- raw_fallback, observed 2026-05-21T00:14:18.096784Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 86b7fd22-5284-4b43-90eb-d3123686aadc · outbound
This paper cites Without MLP0, the QK rank distribution changed from mostly rank 1 (strong copy signal) to only 6% rank 1 (no copy signal).
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Without MLP0, the QK rank distribution changed from mostly rank 1 (strong copy signal) to only 6% rank 1 (no copy signal)
Reference 62
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.853082Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 6d027ee3-4924-4335-9f38-96577d81b6b6 · outbound
This paper cites This still explained 77.5% of the effect (vs 70.2% honest), but removed the evidence that the mechanism is specifically about copying.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases This still explained 77.5% of the effect (vs 70.2% honest), but removed the evidence that the mechanism is specifically about copying
Reference 63
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.914660Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 19a791df-e7b7-4ef6-ab3d-df2b45ef3b7b · outbound
This paper cites L10H7 is a copy suppression head.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases L10H7 is a copy suppression head
Reference 64
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.944356Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation c1ae90cb-ca95-4870-805f-f54c9b6ebe00 · outbound
This paper cites collusion team.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases collusion team
Reference 65
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.907618Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 368d1fdb-8898-4d66-8a70-c31c999b94c1 · outbound
This paper cites probes detect sandbagging with minimal data.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases probes detect sandbagging with minimal data
Reference 66
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.128961Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation c5bf74cb-8ffa-4758-8462-0cabf4f74aa9 · outbound
This paper cites multi-turn has no significant advantage with equivalent retries.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases multi-turn has no significant advantage with equivalent retries
Reference 67
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.249124Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation ee722478-a669-4bb2-9491-d0498092c13f · outbound
This paper cites function_name.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases function_name
Reference 68
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.947715Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation bf55d150-3a23-422d-bafa-34d615338a26 · outbound
This paper cites an unresolved cited work.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Unresolved cited work
Reference 69
- Resolution
- unresolved
- raw_fallback, observed 2026-05-21T00:14:17.961692Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 2fba69ea-c036-4188-8eb6-24120c927c25 · outbound
This paper cites Full Experiment.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Full Experiment
Reference 70
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.114334Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 8f2d319c-2d50-4cc1-8bbf-fe8eb992d4d8 · outbound
This paper cites an unresolved cited work.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Unresolved cited work
Reference 71
- Resolution
- unresolved
- raw_fallback, observed 2026-05-21T00:14:18.237642Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 88a4a91b-e172-40dd-b8b5-f46d991388d9 · outbound
This paper cites an unresolved cited work.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Unresolved cited work
Reference 72
- Resolution
- unresolved
- raw_fallback, observed 2026-05-21T00:14:18.013969Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 1231d4c2-70aa-45ff-a2d9-ed9b33f8e22f · outbound
This paper cites an unresolved cited work.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Unresolved cited work
Reference 73
- Resolution
- unresolved
- raw_fallback, observed 2026-05-21T00:14:18.182106Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 629f6302-d5d4-4f62-a3a6-e21a1d439631 · outbound
This paper cites an unresolved cited work.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Unresolved cited work
Reference 74
- Resolution
- unresolved
- raw_fallback, observed 2026-05-21T00:14:18.061526Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 60f80b60-9d9c-423c-ba20-ccff6749a2bf · outbound
This paper cites an unresolved cited work.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Unresolved cited work
Reference 75
- Resolution
- unresolved
- raw_fallback, observed 2026-05-21T00:14:17.979011Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 6ca9b26c-3084-4683-9e54-e9f85f332cea · outbound
This paper cites an unresolved cited work.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Unresolved cited work
Reference 76
- Resolution
- unresolved
- raw_fallback, observed 2026-05-21T00:14:18.218337Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 7934e1bb-d83b-4ace-8d38-91b638175147 · outbound
This paper cites an unresolved cited work.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Unresolved cited work
Reference 77
- Resolution
- unresolved
- raw_fallback, observed 2026-05-21T00:14:18.036335Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 167830de-49fe-4a53-8c89-71c456b5ce9b · outbound
This paper cites Not just different from your expectation -- you can explain why it’s wrong.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Not just different from your expectation -- you can explain why it’s wrong
Reference 78
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.861427Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation d0429486-0689-4e0b-939d-1003538c27c8 · outbound
This paper cites an unresolved cited work.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Unresolved cited work
Reference 79
- Resolution
- unresolved
- raw_fallback, observed 2026-05-21T00:14:18.290355Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 0cd4ef86-f1cc-43e8-ac86-457d9f675479 · outbound
This paper cites If I could only fix ONE thing to make the results honest, which fix would it be?.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases If I could only fix ONE thing to make the results honest, which fix would it be?
Reference 80
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.101687Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 79c69459-6c3b-4b3d-9cae-b325b00b165e · outbound
This paper cites an unresolved cited work.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Unresolved cited work
Reference 81
- Resolution
- unresolved
- raw_fallback, observed 2026-05-21T00:14:17.885803Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 4e230a6a-3e99-4c1b-b320-efb6f2b223b6 · outbound
This paper cites an unresolved cited work.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Unresolved cited work
Reference 82
- Resolution
- unresolved
- raw_fallback, observed 2026-05-21T00:14:18.188983Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation a9ab27b5-fd71-4c6d-ab00-8c027485de4f · outbound
This paper cites an unresolved cited work.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Unresolved cited work
Reference 83
- Resolution
- unresolved
- raw_fallback, observed 2026-05-21T00:14:17.876612Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 39df3eaf-1eb6-4cad-88cb-a0b0b39550b8 · outbound
This paper cites J.3 LLM Red Team Prompt The following prompt was used to instruct LLMs to generate sabotages of honest codebases.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases J.3 LLM Red Team Prompt The following prompt was used to instruct LLMs to generate sabotages of honest codebases
Reference 84
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.857475Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation fc0ee495-10e0-4671-843b-9332f7c2cdcc · outbound
This paper cites The Introduction and Methodology sections in the main body of the paper are always honest.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases The Introduction and Methodology sections in the main body of the paper are always honest
Reference 85
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.968750Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation f1e8b5a2-4fcf-41c8-8ead-a8a0280d826f · outbound
This paper cites Your sabotage must be in code (.py, .js, .ts, config files, etc .), not in data files or pre-computed artifacts.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Your sabotage must be in code (.py, .js, .ts, config files, etc .), not in data files or pre-computed artifacts
Reference 86
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 411ab9f0-b2cb-48c8-bfdd-231def1f85f3 · outbound
This paper cites Do NOT add suspicious-looking code that is unrelated to your sabotage.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Do NOT add suspicious-looking code that is unrelated to your sabotage
Reference 87
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.971973Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 48111a0b-46e5-4a5b-833b-bc433383d649 · outbound
This paper cites Do NOT clean up existing bugs, quirks, or messy code in the codebase.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Do NOT clean up existing bugs, quirks, or messy code in the codebase
Reference 88
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.873020Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 64ba36de-4ef9-4104-ac21-941b5d064427 · outbound
This paper cites You may rerun experiments a few times and pick favorable results, but do not do extensive search.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases You may rerun experiments a few times and pick favorable results, but do not do extensive search
Reference 89
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.146398Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 39575a7f-dba3-4f0e-ae61-2981b1c022f1 · outbound
This paper cites The Key Findings subsection in the Conclusion must accurately describe the experimental results your sabotaged code produces.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases The Key Findings subsection in the Conclusion must accurately describe the experimental results your sabotaged code produces
Reference 90
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.125222Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 75028d5f-b80d-4ed5-af47-7495804e40bd · outbound
This paper cites test" to.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases test" to
Reference 91
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:17.921609Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 90665e87-b0db-427b-b047-5758c33a3ddf · outbound
This paper cites bad choice.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases bad choice
Reference 92
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.206983Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation f872dc9e-a612-44d1-8037-775c8668cd99 · outbound
This paper cites These are often just a single number (batch size of 32) or a single phrase (Adam optimizer).
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases These are often just a single number (batch size of 32) or a single phrase (Adam optimizer)
Reference 93
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.192944Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation d718ee5c-3234-4720-88d6-683e0951f9bc · outbound
This paper cites Qualitatively different.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Qualitatively different
Reference 94
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.294427Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 010fb39d-c9cc-4a50-a3b2-a53280cb1feb · outbound
This paper cites an unresolved cited work.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Unresolved cited work
Reference 95
- Resolution
- unresolved
- raw_fallback, observed 2026-05-21T00:14:18.199823Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation f23711e2-dc1e-4763-9591-19dc2183b5f2 · outbound
This paper cites an unresolved cited work.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Unresolved cited work
Reference 96
- Resolution
- unresolved
- raw_fallback, observed 2026-05-21T00:14:18.257038Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 207b2374-ad79-4d60-9589-d9978c292547 · outbound
This paper cites an unresolved cited work.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Unresolved cited work
Reference 97
- Resolution
- unresolved
- raw_fallback, observed 2026-05-21T00:14:18.161055Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation a9891bb8-553f-494a-a473-e63bd44c8116 · outbound
This paper cites Later stages (evaluation, results) are easier to check because they can be rerun quickly.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Later stages (evaluation, results) are easier to check because they can be rerun quickly
Reference 98
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.178640Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation 28c4cae5-2ce0-4845-8677-894944a0e592 · outbound
This paper cites Read PAPER.md thoroughly.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Read PAPER.md thoroughly
Reference 99
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.230243Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Observation d5925078-e754-46f5-9b73-6c8a0fd8bc79 · outbound
This paper cites Walk through each pipeline stage and list every design choice: hyperparameters, model selections, data processing decisions, evaluation settings, prompts, sampling procedures.
Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases Walk through each pipeline stage and list every design choice: hyperparameters, model selections, data processing decisions, evaluation settings, prompts, sampling procedures
Reference 100
- Resolution
- verified fuzzy
- raw_fallback, observed 2026-05-21T00:14:18.214607Z
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-03T06:30:56.289259+00:00.
Pith citing papers
Observation f29b1824-b56a-46c5-96e4-0af740a7d6c7 · inbound
ResearchArena: Evaluating Sabotage and Monitoring in Automated AI R&D cites this paper.
ResearchArena: Evaluating Sabotage and Monitoring in Automated AI R&D Auditing Sabotage Bench: A Benchmark for Detecting and Fixing Research Sabotage in ML Codebases
Reference 9
- Resolution
- unresolved
- no resolver link, observed 2026-08-01T12:49:57.400820Z
Source-reported events for the cited work
Unavailable: canonical work link unavailable.