The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance | Paper Citation Record

Observation 4a0b3aa5-66dd-480f-8755-8c8ed4d79bde · outbound

This paper cites American Psychologist , volume=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance American Psychologist , volume=

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.062361Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3dd46683-4846-42e5-98c0-5c1d46a7b828 · outbound

This paper cites Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics , pages=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics , pages=

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.066230Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b9642f4d-e66c-4113-b314-f253f07d72ea · outbound

This paper cites Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing , pages=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing , pages=

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.073148Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 109456f6-e665-4720-9b2f-d386afaa6f09 · outbound

This paper cites A Dataset for Answering Time-Sensitive Questions.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance A Dataset for Answering Time-Sensitive Questions

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.076592Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 24f16c8e-c5b0-4c23-aa29-5710360be67a · outbound

This paper cites Holistic Evaluation of Language Models.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Holistic Evaluation of Language Models

Reference 6

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.080134Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 06e35ed0-ecf8-4fa0-9da9-44fa5be352f6 · outbound

This paper cites Psychometrika , volume=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Psychometrika , volume=

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.083429Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1c3c8517-7422-43d5-9a48-0ee24479e919 · outbound

This paper cites Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics , pages=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics , pages=

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.087044Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ed6785fc-6d2c-49ec-b036-947f4f27296f · outbound

This paper cites IEEE Access , volume=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance IEEE Access , volume=

Reference 9

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.090158Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c93bfb25-5b3e-48ae-9732-18aef02ab399 · outbound

This paper cites Transformed-Linear Models for Time Series Extremes.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Transformed-Linear Models for Time Series Extremes

Reference 10

Resolution: metadata mismatch; local_arxiv, observed 2026-08-16T00:39:52.697193Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 5f8c32a0-f5e8-4940-9a44-6dffdd8e65f1 · outbound

This paper cites ACM Transactions on Intelligent Systems and Technology , year=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance ACM Transactions on Intelligent Systems and Technology , year=

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.096901Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation fb2c94a3-3d22-4283-940d-7ec4970dcae0 · outbound

This paper cites Proceedings of ACL , pages=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Proceedings of ACL , pages=

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.100276Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7f34a779-b703-44c9-9ee9-c204200b007c · outbound

This paper cites Ignore Previous Prompt: Attack Techniques For Language Models.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Ignore Previous Prompt: Attack Techniques For Language Models

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.103724Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a368030c-3898-4c62-9d52-31ef4d66fad4 · outbound

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

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Universal and Transferable Adversarial Attacks on Aligned Language Models

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.107243Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 072b0d71-7e86-4954-8c51-6d0ff14ca349 · outbound

This paper cites Trusting Your Evidence: Hallucinate Less with Context-aware Decoding.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Trusting Your Evidence: Hallucinate Less with Context-aware Decoding

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.117507Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f7fbf4f4-ac29-4026-be64-e95332c41bbb · outbound

This paper cites Distill , year=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Distill , year=

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.128030Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9a2fa57f-161e-4465-aaee-5312862ce30c · outbound

This paper cites ACM Computing Surveys , volume=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance ACM Computing Surveys , volume=

Reference 21

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:53.050672Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 672faa8e-4a09-4dd5-89ff-6203d73766f2 · outbound

This paper cites arXiv preprint arXiv:2401.xxxxx , year=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance arXiv preprint arXiv:2401.xxxxx , year=

Reference 22

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:53.040611Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 8d6663f4-c8f0-4476-a826-29a807ccb08f · outbound

This paper cites Debug like a Human: A Large Language Model Debugger via Verifying Runtime Execution Step-by-step.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Debug like a Human: A Large Language Model Debugger via Verifying Runtime Execution Step-by-step

Reference 24

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.142234Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cac133d0-4b3c-4ea3-9cd9-07445236a2f9 · outbound

This paper cites arXiv preprint arXiv:2312.xxxxx , year=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance arXiv preprint arXiv:2312.xxxxx , year=

Reference 25

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:53.031352Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation a473920b-5e73-46bc-99f0-d9ab77f7ac1d · outbound

This paper cites Teaching Large Language Models to Self-Debug.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Teaching Large Language Models to Self-Debug

Reference 26

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.149556Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b01fae02-8e58-4b60-b5a3-ceea22c4625c · outbound

This paper cites Replacing Judges with Juries: Evaluating LLM Generations with a Panel of Diverse Models.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Replacing Judges with Juries: Evaluating LLM Generations with a Panel of Diverse Models

Reference 30

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.164587Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2d59c4a8-b4fc-40b4-b858-d269151199fd · outbound

This paper cites Proceedings of ICLR , year=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Proceedings of ICLR , year=

Reference 31

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.167887Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c93766a7-0435-4550-b113-c690e6875793 · outbound

This paper cites Advances in Neural Information Processing Systems , pages=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Advances in Neural Information Processing Systems , pages=

Reference 33

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:53.018148Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 83042d05-e2c3-4d49-8890-876883ea9f6d · outbound

This paper cites Least-to-Most Prompting Enables Complex Reasoning in Large Language Models.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Least-to-Most Prompting Enables Complex Reasoning in Large Language Models

Reference 34

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.177206Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation fe72cc17-d14f-4721-900a-fc0fabb284f0 · outbound

This paper cites Measuring and Narrowing the Compositionality Gap in Language Models.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Measuring and Narrowing the Compositionality Gap in Language Models

Reference 36

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.183790Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4225dfae-6b2c-4c79-bc4e-fcd2b5abc957 · outbound

This paper cites Proceedings of ICLR , year=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Proceedings of ICLR , year=

Reference 37

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:53.009243Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 820012fd-dd33-4268-abda-d175a099ce9c · outbound

This paper cites Proceedings of ICLR , year=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Proceedings of ICLR , year=

Reference 38

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:53.000346Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 01b1d51f-2108-4d1d-854d-f19f74a9e283 · outbound

This paper cites Large Language Models Are Human-Level Prompt Engineers.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Large Language Models Are Human-Level Prompt Engineers

Reference 39

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.193011Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 29845126-0a0a-4e85-8fa6-4a00d45bfe51 · outbound

This paper cites Query Rewriting for Retrieval-Augmented Large Language Models.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Query Rewriting for Retrieval-Augmented Large Language Models

Reference 41

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.199842Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a6422781-e41a-43bf-bee4-6ee22d1a9e28 · outbound

This paper cites Proceedings of NAACL-HLT , pages=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Proceedings of NAACL-HLT , pages=

Reference 42

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:52.991815Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 11bd3b0f-0322-4c0e-b817-32c0a374db68 · outbound

This paper cites arXiv preprint arXiv:2412.xxxxx , year=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance arXiv preprint arXiv:2412.xxxxx , year=

Reference 43

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:52.983407Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation ded91785-5ec3-42e9-bc8f-5cea2ac935cd · outbound

This paper cites arXiv preprint arXiv:2412.xxxxx , year=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance arXiv preprint arXiv:2412.xxxxx , year=

Reference 45

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:52.974864Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 000d0a28-e75c-464e-896e-de176bd66c77 · outbound

This paper cites arXiv preprint arXiv:2412.xxxxx , year=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance arXiv preprint arXiv:2412.xxxxx , year=

Reference 46

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:52.964651Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 3542f226-4608-4309-85d5-dc5b1ddf4658 · outbound

This paper cites arXiv preprint arXiv:2405.xxxxx , year=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance arXiv preprint arXiv:2405.xxxxx , year=

Reference 47

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:52.956264Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 8d6eea9f-4b1e-4593-a1ab-58dab6b12eda · outbound

This paper cites 2024 , note=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance 2024 , note=

Reference 48

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:52.946965Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation cd6ceb40-3b06-41fe-881a-9e6245e6b39e · outbound

This paper cites 2024 , note=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance 2024 , note=

Reference 49

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:52.937812Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 57438c98-649d-487b-b6e7-142945df7d2d · outbound

This paper cites an unresolved cited work.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Unresolved cited work

Reference 51

Resolution: unresolved; raw_fallback, observed 2026-08-16T00:39:52.928962Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 7d29956f-a067-4240-8eca-c66d90f06a0d · outbound

This paper cites an unresolved cited work.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Unresolved cited work

Reference 52

Resolution: unresolved; raw_fallback, observed 2026-08-16T00:39:52.919566Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation e0aa3ae6-2b91-4268-b819-c223936dd16c · outbound

This paper cites an unresolved cited work.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Unresolved cited work

Reference 53

Resolution: unresolved; raw_fallback, observed 2026-08-16T00:39:52.910724Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation cd38b232-9403-4df0-bb7b-ba0920572adb · outbound

This paper cites an unresolved cited work.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Unresolved cited work

Reference 54

Resolution: unresolved; raw_fallback, observed 2026-08-16T00:39:52.900886Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation c6baaf8a-4275-4730-a995-b4ff038e2a11 · outbound

This paper cites an unresolved cited work.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Unresolved cited work

Reference 55

Resolution: unresolved; raw_fallback, observed 2026-08-16T00:39:52.891221Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 4f7d71ac-6318-4ab3-b766-50d9f2c849a5 · outbound

This paper cites , title =.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance , title =

Reference 56

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:52.882272Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation b86b5a10-00a1-4759-b229-ce0607d566ea · outbound

This paper cites 2025 , eprint=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance 2025 , eprint=

Reference 57

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.249521Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3b18db8e-9d6d-4c70-8581-153af402b92c · outbound

This paper cites an unresolved cited work.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Unresolved cited work

Reference 58

Resolution: unresolved; raw_fallback, observed 2026-08-16T00:39:52.867977Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 22e85d94-484f-46fd-9483-586f0a93cb95 · outbound

This paper cites 2025 , eprint=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance 2025 , eprint=

Reference 59

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:52.859486Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 2ca7759e-e40d-4ae5-8c3a-97556c259832 · outbound

This paper cites 2024 , eprint=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance 2024 , eprint=

Reference 60

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.259041Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8a6e6c99-5829-4ce1-9e8b-ef6f7ac7fabb · outbound

This paper cites 2024 , eprint=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance 2024 , eprint=

Reference 61

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.262447Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f9c02850-0ce2-4ba5-96c2-1194fcceb87c · outbound

This paper cites 2024 , eprint=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance 2024 , eprint=

Reference 62

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.265510Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 339028c1-69e8-4f57-8df2-8af0a7e2d536 · outbound

This paper cites 2024 , eprint=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance 2024 , eprint=

Reference 63

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.268216Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5b055ac8-3d62-41c1-8833-ede914938f82 · outbound

This paper cites 2024 , institution=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance 2024 , institution=

Reference 64

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:52.832136Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation b17afc4a-b8b4-4f7e-aa94-6d2d817896ad · outbound

This paper cites 2024 , url=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance 2024 , url=

Reference 65

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:52.822731Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 10bcb47f-3c67-414c-9560-eaa8737641ac · outbound

This paper cites 2024 , eprint=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance 2024 , eprint=

Reference 66

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:52.813623Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 3cdcf9d4-da17-4dd1-816d-721de9e4eac1 · outbound

This paper cites 2024 , eprint =.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance 2024 , eprint =

Reference 67

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.280836Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8bf043df-0fea-4a38-9cee-46ccd62e5ac3 · outbound

This paper cites 2021 , eprint =.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance 2021 , eprint =

Reference 69

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:52.799689Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation a7c44720-1f06-4fb7-81a3-b3da376fb2e1 · outbound

This paper cites 2023 , eprint =.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance 2023 , eprint =

Reference 71

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.294509Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 60db8979-57cb-42be-af4a-953372c1a009 · outbound

This paper cites Pragmatics in the Era of Large Language Models: A Survey on Datasets, Evaluation, Opportunities and Challenges.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Pragmatics in the Era of Large Language Models: A Survey on Datasets, Evaluation, Opportunities and Challenges

Reference 73

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.300504Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b4695234-aac8-4034-beeb-46d743bf556b · outbound

This paper cites Ant \`o nia and Salam \'o , Maria.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Ant \`o nia and Salam \'o , Maria

Reference 74

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:52.784835Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 7fb97a12-64e1-4773-bae9-942dd1e01df9 · outbound

This paper cites Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics: Industry Track , year=.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics: Industry Track , year=

Reference 77

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:39:52.775938Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 8d3f84f2-1aa7-4dbd-b118-a019b77fed43 · outbound

This paper cites Phi-4 Technical Report.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Phi-4 Technical Report

Reference 78

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.316382Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ff4aadb3-4b63-40cc-a4f4-78639bb99f87 · outbound

This paper cites gpt-oss-120b & gpt-oss-20b Model Card.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance gpt-oss-120b & gpt-oss-20b Model Card

Reference 79

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.319984Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9c1837ae-7470-418e-b754-f645caeeef64 · outbound

This paper cites GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning

Reference 80

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.323512Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 68549af3-e057-437b-bfa1-aaa772e8e4ae · outbound

This paper cites Training Verifiers to Solve Math Word Problems.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Training Verifiers to Solve Math Word Problems

Reference 81

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.326186Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ca9af2cf-4364-47ce-9036-dc4974ec3244 · outbound

This paper cites an unresolved cited work.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Unresolved cited work

Reference 82

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.328984Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bed39e8d-e445-4588-995a-f50c528c54c3 · outbound

This paper cites Robustness Gym: Unifying the NLP Evaluation Landscape.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Robustness Gym: Unifying the NLP Evaluation Landscape

Reference 83

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.331692Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 97170ab4-e3fa-4443-a1b0-a44b2c79de52 · outbound

This paper cites an unresolved cited work.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Unresolved cited work

Reference 84

Resolution: unresolved; raw_fallback, observed 2026-08-16T00:39:52.765947Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 5a984a37-5c10-4b45-b82d-28347514f2a5 · outbound

This paper cites The Llama 3 Herd of Models.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance The Llama 3 Herd of Models

Reference 85

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.337868Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 18949cbc-fe90-4c22-901a-1cd6c46ebb80 · outbound

This paper cites an unresolved cited work.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Unresolved cited work

Reference 86

Resolution: unresolved; raw_fallback, observed 2026-08-16T00:39:52.757017Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation d1f3b4a0-757d-4aca-8055-e3291e2cc6ea · outbound

This paper cites Mistral 7B.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Mistral 7B

Reference 87

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.344190Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bc503bfa-b9b5-4cc7-bf20-09f759493e69 · outbound

This paper cites LLM-Blender: Ensembling Large Language Models with Pairwise Ranking and Generative Fusion.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance LLM-Blender: Ensembling Large Language Models with Pairwise Ranking and Generative Fusion

Reference 88

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.347392Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c7ff56bf-1cb0-4690-a76c-392a433eb579 · outbound

This paper cites DSPy: Compiling Declarative Language Model Calls into Self-Improving Pipelines.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance DSPy: Compiling Declarative Language Model Calls into Self-Improving Pipelines

Reference 89

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.350464Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3bf73bf2-5f24-446f-9077-3cee862f65d8 · outbound

This paper cites Decomposed Prompting: A Modular Approach for Solving Complex Tasks.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Decomposed Prompting: A Modular Approach for Solving Complex Tasks

Reference 90

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.353992Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 29c3c036-d30f-4a0f-9abe-30b683d33de3 · outbound

This paper cites Let's Verify Step by Step.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Let's Verify Step by Step

Reference 91

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.356996Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b01399e6-1792-4743-9e71-5c8ba7ac22bf · outbound

This paper cites Self-Refine: Iterative Refinement with Self-Feedback.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Self-Refine: Iterative Refinement with Self-Feedback

Reference 92

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.359807Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ab384b88-11a8-432f-8c17-4315a86e9ecb · outbound

This paper cites Granite Code Models: A Family of Open Foundation Models for Code Intelligence.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Granite Code Models: A Family of Open Foundation Models for Code Intelligence

Reference 93

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.362576Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2eb428a1-3306-46a2-bb84-3358713eac60 · outbound

This paper cites an unresolved cited work.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Unresolved cited work

Reference 94

Resolution: unresolved; raw_fallback, observed 2026-08-16T00:39:52.747827Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 71c17b2b-a50e-4d4a-9cdd-a44fb3bd3a75 · outbound

This paper cites Automatic Prompt Optimization with "Gradient Descent" and Beam Search.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Automatic Prompt Optimization with "Gradient Descent" and Beam Search

Reference 95

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.369060Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7ab81be8-bf9b-4928-af00-1c1a5236d669 · outbound

This paper cites an unresolved cited work.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Unresolved cited work

Reference 96

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.372551Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c21fd6f6-b1c2-40a2-b2d3-69f9b15646be · outbound

This paper cites an unresolved cited work.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Unresolved cited work

Reference 97

Resolution: unresolved; raw_fallback, observed 2026-08-16T00:39:52.739113Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 8b4eefdf-1dc6-4b40-9cf9-7a367ab500e4 · outbound

This paper cites Quantifying Language Models' Sensitivity to Spurious Features in Prompt Design or: How I learned to start worrying about prompt formatting.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Quantifying Language Models' Sensitivity to Spurious Features in Prompt Design or: How I learned to start worrying about prompt formatting

Reference 98

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.378889Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 18b6ef97-f982-4bfd-b88a-923db5ffc3b4 · outbound

This paper cites Reflexion: Language Agents with Verbal Reinforcement Learning.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Reflexion: Language Agents with Verbal Reinforcement Learning

Reference 99

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.382593Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f4c2d64f-7e56-40b6-b502-cf4246f02e95 · outbound

This paper cites Qwen2.5 Technical Report.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Qwen2.5 Technical Report

Reference 100

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.386246Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4233beed-5a8d-465d-aec7-490d776d96ae · outbound

This paper cites DebugBench: Evaluating Debugging Capability of Large Language Models.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance DebugBench: Evaluating Debugging Capability of Large Language Models

Reference 101

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.390073Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation eabaaace-4c0d-410a-a61b-39be8a7844e7 · outbound

This paper cites Universal Adversarial Triggers for Attacking and Analyzing NLP.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Universal Adversarial Triggers for Attacking and Analyzing NLP

Reference 102

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.393578Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b26d49e1-fb4e-4f5f-aea1-c3d37a80e3a5 · outbound

This paper cites an unresolved cited work.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Unresolved cited work

Reference 103

Resolution: unresolved; raw_fallback, observed 2026-08-16T00:39:52.729732Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 3916d912-f60c-42a2-8593-599f8ac5ae9c · outbound

This paper cites Self-Consistency Improves Chain of Thought Reasoning in Language Models.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Self-Consistency Improves Chain of Thought Reasoning in Language Models

Reference 104

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.399322Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ab7e5815-0df3-4a43-838e-6c7ab4076f46 · outbound

This paper cites Chain-of-Thought Prompting Elicits Reasoning in Large Language Models.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Chain-of-Thought Prompting Elicits Reasoning in Large Language Models

Reference 105

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.402005Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 15a16d15-6928-49d0-b0b1-7c9427d2aeb4 · outbound

This paper cites Qwen3 Technical Report.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Qwen3 Technical Report

Reference 106

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.404876Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 05d3f8d6-143b-477b-a4d8-26ab71841ca2 · outbound

This paper cites Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena.

The Wording Effect: Quantifying Two-Way Drift in LLM Benchmark Performance Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena

Reference 107

Resolution: unresolved; no resolver link, observed 2026-08-16T00:39:52.407918Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Coverage vector

Reference resolution

Outbound references

Pith citing papers