The DOI matches its Crossref record; 3 of the 3 references it cites were found in a registry or an index.
LM-Infinite: Zero-Shot Extreme Length Generalization for Large Language Models
Open accessAbstract
Chi Han, Qifan Wang, Hao Peng, Wenhan Xiong, Yu Chen, Heng Ji, Sinong Wang. Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2024.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Learnable parameters at the start of the sequence can mitigate the causal-mask bias by redistributing attention weights more effectively.
“these learnable parameters at the start of the sequence can mitigate such effect by redistributing the attention weights more effectively (Han et al. 2024; Xiao et al. 2024).”
From Titans: Learning to Memorize at Test Time (2025) · cites this paper as Han et al. 2024
The available passages contain only the title and bibliographic metadata, with no substantive section or page text. Searches for learnable parameters, initial tokens, causal-mask bias, and attention-weight redistribution found no passage addressing the claim; the decisive text may be missing from our copy. (LLM, conf=0.70)
Experimental — accuracy not yet measuredAttention with causal mask has implicit bias toward initial tokens in the sequence, and so attention weights are almost always highly active for initial tokens, resulting in performance damage.
“Attention with causal mask has implicit bias toward initial tokens in the sequence, and so attention weights are almost always highly active for initial tokens, resulting in performance damage.”
From Titans: Learning to Memorize at Test Time (2025) · cites this paper as Han et al. 2024
The supplied passages contain only bibliographic metadata (no section or page), not discussion of attention behavior. Searches for causal masks, initial-token attention, attention sinks, and performance effects returned no relevant passage from our copy, so the claim cannot be verified from it. (LLM, conf=0.73)
Experimental — accuracy not yet measuredExisting KV cache eviction methods typically retain KV pairs with high attention weights but overlook the impact of attention redistribution caused by token removal and spatial-temporal dynamics in KV selection.
“Existing eviction methods typically retain KV pairs with high attention weights but overlook the impact of attention redistribution caused by token removal, as well as the spatial-temporal dynamics in KV selection.”
From ReST-KV: Robust KV Cache Eviction with Layer-wise Output Reconstruction and Spatial-Temporal Smoothing (2026) · cites this paper as Han et al., 2024
The available copy provides an abstract and bibliographic metadata, but no passage addresses how KV-cache eviction methods select pairs. Searches for attention redistribution, token removal, and dynamic KV selection found no passage supporting or contradicting the claim. (LLM, conf=0.72)
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
LLaMA: Open and Efficient Foundation Language Models
Touvron, H., Lavril, T., Izacard, G., Martinet, X., Lachaux, M.-A., Lacroix, T., Rozière, B., Goyal, N., Hambro, E., Azhar, F., et al. (2023) · arXiv · doi:10.48550/arxiv.2302.13971 · Cited by 4,602
RoFormer: Enhanced transformer with rotary position embedding
Su, J., Ahmed, M., Lu, Y., Pan, S., Bo, W., & Liu, Y. (2024) · Neurocomputing · doi:10.1016/j.neucom.2023.127063 · Cited by 1,957
LongLLMLingua: Accelerating and Enhancing LLMs in Long Context Scenarios via Prompt Compression
Jiang, H., Wu, Q., Luo, X., Li, D., Lin, C.-Y., Yang, Y., & Qiu, L. (2024) · Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · doi:10.18653/v1/2024.acl-long.91 · Cited by 400
Flagged first, then by citations
Research works printed but not linked: 4
- LM-Infinite: Zero-Shot Extreme Length Generalization for Large Language Models https://aclanthology.org/2024.naacl-long.222/ (Han et al., NAACL 2024)
- LM-Infinite: Zero-Shot Extreme Length Generalization for Large Language Models (Han et al., NAACL 2024)
- Chi Han, Qifan Wang, Hao Peng, Wenhan Xiong, Yu Chen, Heng Ji, and Sinong Wang. 2024. LM-Infinite: Zero-Shot Extreme Length Generalization for Large Language Models. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers), pages 3991–4008, Mexico City, Mexico. Association for Computational Linguistics.
- ACL materials are Copyright © 1963–2026 ACL; other materials are copyrighted by their respective copyright holders. Materials prior to 2016 here are licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 International License. Permission is granted to make copies for the purposes of teaching and research. Materials published in or after 2016 are licensed on a Creative Commons Attribution 4.0 International License.
Cited by
Sorted by citations
RULER: What's the Real Context Size of Your Long-Context Language Models?
Hsieh, C.-P., Sun, S., Kriman, S., Acharya, S., Rekesh, D., Jia, F., Zhang, Y., & Ginsburg, B. (2024) · arXiv · Cited by 1,068
Titans: Learning to Memorize at Test Time
Behrouz, A., Zhong, P., & Mirrokni, V. (2025) · arXiv · doi:10.48550/arxiv.2501.00663 · Cited by 24
“From the technical perspective, these learnable parameters at the start of the sequence can mitigate such effect by redistributing the attention weights more effectively [120, 121].”
Transformers are Multi-State RNNs
Oren, M., Hassid, M., Yarden, N., Adi, Y., & Schwartz, R. (2024) · arXiv · doi:10.48550/arxiv.2401.06104 · Cited by 16
FINCH: Prompt-guided Key-Value Cache Compression for Large Language Models
Corallo, G., & Papotti, P. (2024) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00716 · Cited by 5
ChunkKV: Semantic-Preserving KV Cache Compression for Efficient Long-Context LLM Inference
Liu, X., Tang, Z., Dong, P., Li, Z., Liu, Y., Li, B., Hu, X., & Chu, X. (2025) · arXiv · doi:10.48550/arxiv.2502.00299 · Cited by 4
Rethinking Memory in LLM based Agents: Representations, Operations, and Emerging Topics
Du, Y., Huang, W., Zheng, D., Wang, Z., Montella, S., Lapata, M., Wong, K.-F., & Pan, J. Z. (2025) · arXiv · doi:10.48550/arxiv.2505.00675 · Cited by 1
Make Each Token Count: Towards Improving Long-Context Performance with KV Cache Eviction
Bui, N., Nguyen, H. T., Cohan, A., & Ying, R. (2026) · arXiv · doi:10.48550/arxiv.2605.09649
“Solving this combinatorial problem exactly at every decoding step is infeasible, so most prior work relies on heuristic eviction rules [Xiao et al., 2023; Han et al., 2023; Zhang et al., 2023; Li et al., 2024c; Cai et al., 2025; Ghadia et al., 2025].”
An, Y., Lu, C., Zhu, K., Yu, T., Zhao, C., Wu, H., Tang, M., & Wang, J. (2026) · arXiv · doi:10.48550/arxiv.2605.08840
FlexPrefill: A Context-Aware Sparse Attention Mechanism for Efficient Long-Sequence Inference
Lai, X., Lu, J., Luo, Y., Ma, Y., & Zhou, X. (2025) · arXiv · doi:10.48550/arxiv.2502.20766
Metadata
- Authors
- Han, C., Wang, Q., Peng, H., Xiong, W., Chen, Y., Ji, H., & Wang, S.
- Venue
- Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)
- Published
- 2024
- Pages
- 3991-4008
- OpenAlex
- W4401042914
- Cited by
- 22
- References
- 3
- Citations by year
- 2026: 12 · 2025: 8 · 2024: 2
- FWCI
- 6.35
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- Locations
- Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT) (publishedVersion, cc-by)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Physical Sciences › Computer Science › Artificial Intelligence › Natural Language Processing Techniques
- Keywords
- zero-shot generalization, long-form text generation, large language models
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
