Pith. sign in

REVIEW 4 major objections 6 minor 58 references

Neon: News Entity-Interaction Extraction for Enhanced Question Answering

T0 review · 4 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Timestamped entity-interaction tuples from news improve temporal QA and rival live web search.

desk verdict Neon is a well-built temporal QA system with a new real-query dataset, but its headline evaluation relies on an LLM judge that correlates weakly with humans, so the claimed edge over web search is not established. read the letter →

arxiv 2411.12449 v2 pith:XJQLRFHG submitted 2024-11-19 cs.CL cs.IR

Sneha Singhania , Silviu Cucerzan , Allen Herring , Sujay Kumar Jauhar This is my paper · ORCID
classification cs.CLcs.IR
keywords temporalquestionansweringinformationextractionlargelanguagemodelsknowledgegraphretrieval-augmentedgenerationentity-centricquerynewsstreamsopen
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to establish that a timestamped graph of short entity-interaction tuples, extracted from news streams by an LLM, can make question answering about recent events more accurate and grounded than feeding the same model raw news passages. The authors argue that compact, entity-centric propositions avoid the 'lost in the middle' degradation of long contexts and give the LLM exactly the dated facts it needs. If correct, this would offer a lightweight alternative to live web search for entity-centric temporal queries, with results that rival a commercial search engine in few-shot evaluation.

What carries the argument

The load-bearing mechanism is the LLM-generated lexicalized entity-interaction tuple $(t, \text{NEI}(s,o))$ — a single natural-language sentence describing an event or activity between entities, stamped with the article's timestamp. The paper calls these 'news entity-interactions' and stores them in a temporal datastore indexed by dense embeddings. Retrieval either matches the query's timestamp exactly (with a $\pm r$-day back-off) or uses semantic similarity over the reformulated, time-stamped query. Because each tuple is one sentence, the prompt context stays compact, avoiding the lost-in-the-middle and information-overload problems that hurt longer news chunks.

What would settle it

Have two annotators verify a random sample of 200 Neon(M2) tuples against the original news chunks and article timestamps, computing precision and timestamp-accuracy rates; if a substantial share of tuples are unsupported or misdated, the faithfulness gains reported in Tables 4 and 5 would not reflect genuine grounding.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central discovery is that replacing raw news chunks with LLM-generated openIE-style tuples, each stamped with the article's date, makes an LLM answer temporal entity-centric queries more faithfully and, with few-shot evaluation prompts, gives the highest average score across helpfulness, relevance, and faithfulness, ahead of live web search. This is achieved by two graph construction variants: Neon(M1), which extracts interactions of a single subject with marked entities, and Neon(M2), which targets explicit subject-object pairs selected by co-occurrence and TF-IDF, producing a larger, more detailed graph. The paper reports that under temporal retrieval, Neon(M2) outperforms WebRAG on the overall average in the few-shot setting, and that both Neon variants beat news-chunk RAG on helpfulness and faithfulness.

Load-bearing premise

The LLM-written tuples in the Neon graph accurately reflect the news articles they were extracted from, with correct timestamps, even though the paper never measures tuple precision or recall.

Editorial extensions

If this is right

  • A precomputed entity-interaction graph can serve temporal QA without an external search engine, using a retriever-agnostic dense index.
  • Injecting the query date explicitly during reformulation is enough to activate effective temporal retrieval.
  • Doubling the supporting passages from five to ten lowers faithfulness for every method, with the drop more acute for longer passages.
  • A hybrid system that combines Neon's faithfulness with web search's breadth could improve all three metrics, as the paper leaves to future work.
  • The larger Neon(M2) graph, about four times the size of Neon(M1), yields the highest overall average score under few-shot temporal retrieval.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper never measures tuple precision or recall, so the faithfulness scores may partly reflect the LLM's confidence in its own paraphrases rather than verified grounding in the articles.
  • Given the weak correlation between human and automated ratings (0.12–0.24), the absolute Likert scores should be read cautiously even if the relative ordering of methods still holds.
  • The same compact-tuple recipe could transfer to other fast-moving text streams, such as financial filings or sports feeds, with timestamps taken from the stream items.
  • Replacing exact-date matching with a learned time-relevance scoring function could improve retrieval on sparsely covered dates beyond the paper's $\pm r$-day back-off.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper presents Neon, a framework for constructing an entity-centric, timestamped knowledge graph from news streams by prompting LLMs to produce openIE-style interaction tuples (subject, object, relation verbalization). Two graph variants are proposed: Neon(M1), which extracts interactions around a single subject entity, and Neon(M2), which extracts interactions for explicitly paired subject-object entities using TF-IDF-selected co-occurring pairs. The authors evaluate the framework on temporal entity-centric question answering using a dataset of roughly 3,000 real Bing queries across 50 entities and a four-month news corpus. They compare retrieval-augmented generation using Neon tuples against retrieval from news chunks (NewsRAG) and live Bing snippets (WebRAG), with responses generated by GPT-4o and scored by GPT-4o on 3-point Likert scales for helpfulness, relevance, and faithfulness. The main reported result is that, under few-shot evaluation prompts, Neon(M2) achieves the highest overall average score across the three metrics in the temporal retrieval setting, rivaling a live web-scale search engine. The paper also reports a human assessment on 100 responses showing weak positive correlations between human and automatic scores, and it documents evaluation artifacts such as off-scale ratings and a correlation between response length and scores.

Significance. If the central claim held, Neon would make a useful empirical contribution: a compact, structured representation of timestamped entity interactions that can be injected into LLM prompts to improve temporal entity-centric QA, with performance competitive against live web retrieval. The work has genuine strengths: it uses real user queries from search logs, includes both zero-shot and few-shot evaluation settings, reports a human annotation study, and candidly discloses known artifacts of LLM-based evaluation. However, the paper's own evaluation evidence is the major weakness. The correlation between the automated judge and human scores is weak for helpfulness and faithfulness, and essentially zero for relevance (0.12), which is the attribute the method is specifically intended to improve. The paper also never measures the precision, recall, or temporal accuracy of the Neon tuples that constitute the whole system.

major comments (4)
  1. [Section 4, Human Assessment] The paper's headline result is based entirely on the few-shot GPT-4o judge scores used in Tables 4 and 5, yet the Human Assessment subsection reports Pearson correlations between human ratings and these automatic scores of only 0.24 (helpfulness), 0.12 (relevance), and 0.19 (faithfulness) over 100 responses. Relevance is the attribute that temporal retrieval is specifically designed to improve, and its correlation is effectively zero. Consequently, the claim in Table 5 that 'Neon(M2) achieves the highest overall average score across metrics' and the abstract's 'substantial improvements' are not established as claims about QA quality; the apparent advantage could be an artifact of the judge preferring longer, more structured responses. The authors must either report method-by-method comparisons using human scores (or a recalibrated judge validated against humans on the full evaluation set) and provide significance tests, or substantially soften the central claims.
  2. [Section 2.1, Graph Construction] The Neon graph is the core contribution, but its construction is never evaluated. Section 2.1 describes prompting GPT-4o to generate lexicalized NEI tuples and assigning each tuple the timestamp of the source chunk, yet the paper reports no precision, recall, or temporal-accuracy figures for these tuples. If a nontrivial fraction of tuples are hallucinated, ungrammatical, or misattributed to the wrong date, the faithfulness scores in Tables 4 and 5 only measure whether the response uses the provided passages, not whether those passages are true or temporally aligned. The authors should sample Neon tuples and have annotators judge (a) whether the interaction is entailed by the source chunk, (b) whether it is semantically valid, and (c) whether the timestamp matches the event described; the extraction quality should be reported for both Neon(M1) and Neon(M2).
  3. [Section 4, Tables 4 and 5] Tables 4 and 5 report means on a 3-point Likert scale without variance, confidence intervals, or significance tests. The key differences are small—for example, the temporal few-shot average is 1.55 for Neon(M2) versus 1.48 for NewsRAG and 1.52 for WebRAG (generic). These margins are on the same order as the noise implied by the weak human-automation correlations. Without per-query paired tests or bootstrap confidence intervals, the claim that Neon 'rivals' or 'substantially improves' over baselines is not statistically grounded. At minimum, the paper should report standard errors or confidence intervals and a paired significance test for the headline comparisons.
  4. [Sections 2.1, 3, and 4] There is a potential circularity in the evaluation loop: GPT-4o is used to generate the Neon tuples (Section 2.1), to generate the QA responses (Section 3, Evaluation), and to rate those responses on the three Likert attributes (Section 3, Evaluation). The paper does not control for judge-model preferences for outputs that resemble its own style of tuple-flavored, structured text. The response-length analysis in Section 5 and Table A3 suggests exactly such a confounding bias, since longer responses receive higher scores for helpfulness and relevance. Using a different model (e.g., a different family or a fine-tuned evaluator) for the automatic assessment, or reporting human evaluation on the full set, would materially strengthen the claim that the observed gains reflect temporal relevance rather than formatting preferences.
minor comments (6)
  1. [Section 2.1, formal definition] The formal definition G = (E, T, R) does not include the lexicalized interaction descriptions I that are introduced in the same paragraph; clarify whether I is part of the graph or a notational shorthand for the verbalization function.
  2. [Table A4, Prompt templates] The prompt templates for Relevance and Faithfulness contain a copy-paste error: they instruct the model to provide 'reasoning for the helpfulness rating' instead of reasoning for the relevance/faithfulness rating.
  3. [Section 2.2, Temporal Retrieval] The temporal retrieval strategy uses an exact timestamp match on t_d = t_q with a ±r day fallback, but the paper does not specify how 'timestamp' is normalized across different time zones or source publication times; a brief clarification would help reproducibility.
  4. [Section 3, News Articles] The entity linking system NEMO is cited only via a 2014 ERD paper and is described as proprietary; for a research audience, at least a high-level description of its accuracy or recall on the news domain would help calibrate the trust one can place in the entity markup that drives both Neon variants.
  5. [Section 3, Evaluation] The three quality attributes are defined in the text, but their mutual correlations are not examined despite the paper noting that helpfulness and relevance correlate >0.8 in both human and automatic scores; this threatens the independence of the averaged metric and should be discussed.
  6. [General] The paper does not release code, data, or the set of 3,000 queries (likely due to privacy), but it also does not state a clear availability statement; readers are left without a way to reproduce or build on the results.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the Neon QA gains are empirical and comparative, though the LLM-based judge's weak human correlation (relevance 0.12) is a validity risk rather than a circular step.

full rationale

No circular step can be exhibited. Neon is an empirical construction (Section 2.1) and the evaluation (Section 4) compares system outputs on a Likert scale; no equation or parameter fit makes the outcome equal to the input. The only author-overlap citations are NEMO [9] for entity linking and [8] for disambiguation; these are pipeline components, not uniqueness theorems, and the central RAG comparison does not reduce to them. The paper itself documents serious judge-calibration problems: human/automated Pearson correlations of 0.24 (helpfulness), 0.12 (relevance), and 0.19 (faithfulness), off-scale ratings of 5 and 9 on a 3-point scale, and response-length correlations with scores (Table A3). These undermine confidence in the headline 'highest overall average score' (Table 5), but they are measurement-validity risks, not circularity-by-construction. The fact that GPT-4o generated the tuples and also rated the answers is a same-family evaluator concern, not a definitional equivalence. Score 2 reflects one minor self-citation (NEMO) and the self-referential evaluation loop, neither of which forces the result.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The central claim rests on unverified domain assumptions about extraction quality and evaluation validity, plus several unspecified hyperparameters. No code, data, or formal proofs are provided. The Neon graph is a data structure over real entities and events, not a new postulated physical or conceptual entity.

free parameters (6)
  • chunk_size_m
    Number of sentences per news chunk; hyperparameter that controls context detail and LLM prompt size (Section 2.1, step 3), value not reported.
  • batch_size_k
    Number of chunks per prompt for Neon(M2) extraction; hyperparameter that trades cost, context detail, and lost-in-the-middle effects (Section 2.1, M2 step 3).
  • retrieval_top_k = 10 for Neon, 5 for NewsRAG/WebRAG
    Number of retrieved entries per query; varied in Figure 4, chosen to roughly equalize context length, value not justified beyond that.
  • temporal_range_r
    Day-range around query timestamp used to backfill temporal retrieval (Section 2.2), value not reported.
  • spike_threshold = 1 standard deviation above 3-day rolling mean
    Threshold for selecting spiking dates and queries in dataset construction (Section 3).
  • tfidf_pair_threshold
    Criterion for selecting co-occurring subject-object candidate pairs in Neon(M2) (Section 2.1, M2 step 1), not specified.
assumptions (4)
  • domain assumption LLM-generated openIE tuples are accurate and faithful paraphrases of the news chunks.
    Section 2.1 treats the generated NEI(.) tuples as facts for the graph. No precision/recall evaluation of the tuples is provided.
  • domain assumption LLM-based 3-point Likert scores are valid proxies for answer quality.
    Section 3 Evaluation uses LLM judges; the paper's own human assessment shows weak positive correlations (0.12 relevance, 0.19 faithfulness), so this axiom is questionable.
  • domain assumption NEMO entity linking correctly marks all named entities and co-references.
    Section 2.1, step 1 relies on NEMO's XML markup; linking errors would propagate into chunk retrieval and tuple extraction.
  • domain assumption Article publication timestamp equals the time of the events described.
    Section 2.1 assigns each tuple the article's timestamp from the URL. Events may have occurred earlier or be reported late.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Neon: News Entity-Interaction Extraction for Enhanced Question Answering." pith.science (2026). https://pith.science/paper/XJQLRFHG

@misc{pith2026241112449,
  author       = {Pith},
  title        = {Pith review of: Neon: News Entity-Interaction Extraction for Enhanced Question Answering},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XJQLRFHG}},
  note         = {Machine review of arXiv:2411.12449}
}
read the original abstract

Capturing fresh information in near real-time and using it to augment existing large language models (LLMs) is essential to generate up-to-date, grounded, and reliable output. This problem becomes particularly challenging when LLMs are used for informational tasks in rapidly evolving fields, such as Web search related to recent or unfolding events involving entities, where generating temporally relevant responses requires access to up-to-the-hour news sources. However, the information modeled by the parametric memory of LLMs is often outdated, and Web results from prototypical retrieval systems may fail to capture the latest relevant information and struggle to handle conflicting reports in evolving news. To address this challenge, we present the NEON framework, designed to extract emerging entity interactions -- such as events or activities -- as described in news articles. NEON constructs an entity-centric timestamped knowledge graph that captures such interactions, thereby facilitating enhanced QA capabilities related to news events. Our framework innovates by integrating open Information Extraction (openIE) style tuples into LLMs to enable in-context retrieval-augmented generation. This integration demonstrates substantial improvements in QA performance when tackling temporal, entity-centric search queries. Through NEON, LLMs can deliver more accurate, reliable, and up-to-date responses.

Figures

Figures reproduced from arXiv: 2411.12449 by the authors.

Figure 1
Figure 1. Example for entity-centric, time-specific QA. Graph shows search interest for [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Implemented temporal QA pipeline 3 Experimental Setup Entities and Queries. We selected 50 entities by using stratified sampling across four categories—artists, companies, leaders, and pioneers—from Time’s 100 most influential people 2 and Fortune500 companies 3 to ensure a balanced diversity in category, demograph￾ics, and popularity. For each entity, we collected a large volume of user queries from Bing search log… view at source ↗
Figure 3
Figure 3. Coverage for 50 entities in 500 diverse news sources [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Performance comparison with varying top- [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

58 extracted references · 50 canonical work pages

  1. [1]

    Badr AlKhamissi, Millicent Li, Asli Celikyilmaz, Mona Diab, and Marjan Ghazvininejad. 2022. A Review on Language Models as Knowledge Bases. arXiv:2204.06031 [cs.CL]

    arXiv 2022

  2. [2]

    Akari Asai, Zeqiu Wu, Yizhong Wang, Avirup Sil, and Hannaneh Hajishirzi. 2024. Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection. In 12th International Conference on Learning Representations . ICLR, Vienna, Austria

    work page 2024

  3. [3]

    Krisztian Balog. 2018. Entity-Oriented Search

    work page 2018

  4. [4]

    Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, Sandhini Agarwal, Ariel Herbert-Voss, Gretchen Krueger, Tom Henighan, Rewon Child, Aditya Ramesh, Daniel Ziegler, Jeffrey Wu, Clemens Winter, Chris Hesse, Mark Chen, Eric Sigler, Mateusz Litwin, Scott Gr...

    work page 2020

  5. [5]

    Cheng-Han Chiang and Hung-yi Lee. 2023. Can Large Language Models Be an Alternative to Human Evaluations?. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) . ACL, Toronto, 15607–15631

    work page 2023

  6. [6]

    Roi Cohen, Mor Geva, Jonathan Berant, and Amir Globerson. 2023. Crawling The Internal Knowledge-Base of Language Models. In Findings of the Association for Computational Linguistics: EACL 2023 . ACL, Dubrovnik, Croatia, 1856–1869

    work page 2023

  7. [7]

    Together Computer. 2023. RedPajama: an Open Dataset for Training Large Lan- guage Models. Together AI

    work page 2023

  8. [8]

    Silviu Cucerzan. 2007. Large-Scale Named Entity Disambiguation Based on Wikipedia Data. In Proceedings of the 2007 Joint Conference on Empirical Methods in NLP and Computational Natural Language Learning (EMNLP-CoNLL) . ACL, Prague, Czech Republic, 708–716

    work page 2007

Show all 58 references
  1. [9]

    Silviu Cucerzan. 2014. Name entities made obvious: the participation in the ERD 2014 evaluation. In Proceedings of the First International Workshop on Entity Recognition & Disambiguation (Gold Coast, Queensland, Australia). ACM, New York, NY, USA, 95–100

    2014

  2. [10]

    Tim Dettmers, Artidoro Pagnoni, Ari Holtzman, and Luke Zettlemoyer. 2024. QLORA: efficient finetuning of quantized LLMs. In Proceedings of the 37th Inter- national Conference on Neural Information Processing Systems (New Orleans, LA, USA). NeurIPS, NY, USA, 28 pages

    2024

  3. [11]

    Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human...

    2019

  4. [12]

    Darren Edge, Ha Trinh, Newman Cheng, Joshua Bradley, Alex Chao, Apurva Mody, Steven Truitt, and Jonathan Larson. 2024. From Local to Global: A Graph RAG Approach to Query-Focused Summarization. arXiv:2404.16130 [cs.CL]

    2024 arXiv

  5. [13]

    Jiafeng Guo, Gu Xu, Xueqi Cheng, and Hang Li. 2009. Named entity recognition in query. In Proceedings of the 32nd International ACM SIGIR Conference (Boston, MA, USA). ACM, New York, NY, USA, 267–274

    2009

  6. [14]

    Kelvin Guu, Kenton Lee, Zora Tung, Panupong Pasupat, and Ming-Wei Chang

  7. [15]

    Xu Han, Tianyu Gao, Yankai Lin, Hao Peng, Yaoliang Yang, Chaojun Xiao, Zhiyuan Liu, Peng Li, Jie Zhou, and Maosong Sun. 2020. More Data, More Relations, More Context and More Openness: A Review and Outlook for Relation Extraction. In Proceedings of the 1st Conference of the As...

    2020

  8. [16]

    Shibo Hao, Bowen Tan, Kaiwen Tang, Bin Ni, Xiyan Shao, Hengzhe Zhang, Eric Xing, and Zhiting Hu. 2023. BertNet: Harvesting Knowledge Graphs with Arbi- trary Relations from Pretrained Language Models. In Findings of the Association for Computational Linguistics: ACL 2023 . ACL,...

    2023

  9. [17]

    Gautier Izacard, Mathilde Caron, Lucas Hosseini, Sebastian Riedel, Piotr Bo- janowski, Armand Joulin, and Edouard Grave. 2022. Unsupervised Dense Infor- mation Retrieval with Contrastive Learning. TMLR 2022 (2022)

    2022

  10. [18]

    Joel Jang, Seonghyeon Ye, Changho Lee, Sohee Yang, Joongbo Shin, Janghoon Han, Gyeonghun Kim, and Minjoon Seo. 2022. TemporalWiki: A Lifelong Benchmark for Training and Evaluating Ever-Evolving Language Models. In Proceedings of the 2022 Conference on Empirical Methods in NLP ...

    2022

  11. [19]

    Zhen Jia, Philipp Christmann, and Gerhard Weikum. 2024. Faithful Temporal Question Answering over Heterogeneous Sources. In Proceedings of the ACM Web Conference 2024 (Singapore). ACM, New York, NY, USA, 2052–2063

    2024

  12. [20]

    Zhen Jia, Soumajit Pramanik, Rishiraj Saha Roy, and Gerhard Weikum. 2021. Complex Temporal Question Answering on Knowledge Graphs. InProceedings of the 30th ACM International Conference on Information & Knowledge Management (Virtual Event). ACM, New York, NY, USA, 792–802

    2021

  13. [21]

    Xisen Jin, Dejiao Zhang, Henghui Zhu, Wei Xiao, Shang-Wen Li, Xiaokai Wei, Andrew Arnold, and Xiang Ren. 2022. Lifelong Pretraining: Continually Adapting Language Models to Emerging Corpora. In Proceedings of BigScience Episode #5 – Workshop on Challenges & Perspectives in Cre...

    2022

  14. [22]

    Jean Kaddour, Joshua Harris, Maximilian Mozes, Herbie Bradley, Roberta Raileanu, and Robert McHardy. 2023. Challenges and Applications of Large Language Models. arXiv:2307.10169 [cs.CL]

    2023 arXiv

  15. [23]

    Smith, Yejin Choi, and Kentaro Inui

    Jungo Kasai, Keisuke Sakaguchi, Yoichi Takahashi, Ronan Le Bras, Akari Asai, Xinyan Velocity Yu, Dragomir Radev, Noah A. Smith, Yejin Choi, and Kentaro Inui. 2024. REALTIME QA: what’s the answer right now?. In Proceedings of the 37th International Conference on Neural Informat...

    2024

  16. [24]

    Tom Kocmi and Christian Federmann. 2023. Large Language Models Are State- of-the-Art Evaluators of Translation Quality. In Proceedings of the 24th Annual Conference of the European Association for Machine Translation . European Associ- ation for Machine Translation, Tampere, F...

    2023

  17. [25]

    Keshav Kolluru, Vaibhav Adlakha, Samarth Aggarwal, Mausam, and Soumen Chakrabarti. 2020. OpenIE6: Iterative Grid Labeling and Coordination Analysis for Open Information Extraction. In Proceedings of the 2020 Conference on EMNLP . ACL, Online, 3748–3761

    2020

  18. [26]

    Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, and Douwe Kiela. 2020. Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks. In NeurIPS, Vol...

    2020

  19. [27]

    Thomas Lin, Patrick Pantel, Michael Gamon, Anitha Kannan, and Ariel Fuxman

  20. [28]

    Liu, Kevin Lin, John Hewitt, Ashwin Paranjape, Michele Bevilacqua, Fabio Petroni, and Percy Liang

    Nelson F. Liu, Kevin Lin, John Hewitt, Ashwin Paranjape, Michele Bevilacqua, Fabio Petroni, and Percy Liang. 2024. Lost in the Middle: How Language Models Use Long Contexts. TACL 12 (2024), 157–173

    2024

  21. [29]

    Yang Liu, Dan Iter, Yichong Xu, Shuohang Wang, Ruochen Xu, and Chenguang Zhu. 2023. G-Eval: NLG Evaluation using Gpt-4 with Better Human Alignment. In Proceedings of the 2023 Conference on EMNLP . ACL, Singapore, 2511–2522

    2023

  22. [30]

    Kelvin Luu, Daniel Khashabi, Suchin Gururangan, Karishma Mandyam, and Noah A. Smith. 2022. Time Waits for No One! Analysis and Challenges of Temporal Misalignment. In Proceedings of the 2022 Conference of the North Ameri- can Chapter of the Association for Computational Lingui...

    2022

  23. [31]

    Youmi Ma, An Wang, and Naoaki Okazaki. 2023. DREEAM: Guiding Attention with Evidence for Improving Document-Level Relation Extraction. InProceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics. ACL, Dubrovnik, Croatia, 1971–1983

    2023

  24. [32]

    Mounica Maddela, Fernando Alva-Manchego, and Wei Xu. 2021. Controllable Text Simplification with Explicit Paraphrasing. In Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. ACL, Onlin...

    2021

  25. [33]

    Martinez-Rodriguez, Ivan Lopez-Arevalo, and Ana B

    Jose L. Martinez-Rodriguez, Ivan Lopez-Arevalo, and Ana B. Rios-Alvarado. 2018. OpenIE-based approach for Knowledge Graph construction from text. Expert Syst. Appl. 113, C (Dec. 2018), 339–355

    2018

  26. [34]

    Ioannidis, Ades- oji Adeshina, Phillip R

    Costas Mavromatis, Prasanna Lakkur Subramanyam, Vassilis N. Ioannidis, Ades- oji Adeshina, Phillip R. Howard, Tetiana Grinberg, Nagib Hakim, and George Karypis. 2022. TempoQR: Temporal Question Reasoning over Knowledge Graphs. In AAAI-22 Technical Tracks 5. AAAI, virtual, 5825–5833

    2022

  27. [35]

    Costas Mavromatis, Prasanna Lakkur Subramanyam, Vassilis N Ioannidis, Adesoji Adeshina, Phillip R Howard, Tetiana Grinberg, Nagib Hakim, and George Karypis

  28. [36]

    Arvind Neelakantan, Tao Xu, Raul Puri, Alec Radford, Jesse Michael Han, Jerry Tworek, Qiming Yuan, Nikolas Tezak, Jong Wook Kim, Chris Hallacy, Johannes Heidecke, Pranav Shyam, Boris Power, Tyna Eloundou Nekoul, Girish Sastry, Gretchen Krueger, David Schnurr, Felipe Petroski S...

    2022 arXiv

  29. [37]

    Qiang Ning, Hao Wu, Rujun Han, Nanyun Peng, Matt Gardner, and Dan Roth

  30. [38]

    Barlas Oguz, Xilun Chen, Vladimir Karpukhin, Stan Peshterliev, Dmytro Okhonko, Michael Schlichtkrull, Sonal Gupta, Yashar Mehdad, and Scott Yih. 2022. UniK-QA: Unified Representations of Structured and Unstructured Knowledge for Open- Domain Question Answering. In Findings of ...

    2022

  31. [39]

    OpenAI. 2023. GPT-4 Technical Report

    2023

  32. [40]

    Peters, Mark Neumann, Mohit Iyyer, Matt Gardner, Christopher Clark, Kenton Lee, and Luke Zettlemoyer

    Matthew E. Peters, Mark Neumann, Mohit Iyyer, Matt Gardner, Christopher Clark, Kenton Lee, and Luke Zettlemoyer. 2018. Deep Contextualized Word Rep- resentations. In Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguisti...

    2018

  33. [41]

    In Proceedings of the 2020 Conference on EMNLP

    TORQUE: A Reading Comprehension Dataset of Temporal Ordering Ques- tions. In Proceedings of the 2020 Conference on EMNLP . ACL, 1158–1172

    2020

  34. [42]

    Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, and Peter J. Liu. 2020. Limits of Transfer Learning with a Unified Text-to-Text Transformer.JML 21, 140 (2020), 1–67

    2020

  35. [43]

    Ori Ram, Yoav Levine, Itay Dalmedigos, Dor Muhlgay, Amnon Shashua, Kevin Leyton-Brown, and Yoav Shoham. 2023. In-Context Retrieval-Augmented Lan- guage Models. TACL 11 (2023), 1316–1331

    2023

  36. [44]

    Uma Sawant and Soumen Chakrabarti. 2013. Learning joint query interpretation and response ranking. In Proceedings of the 22nd International Conference on World Wide Web (Rio de Janeiro, Brazil). ACM, New York, NY, USA, 1099–1110

    2013

  37. [45]

    Fabio Petroni, Tim Rocktäschel, Sebastian Riedel, Patrick Lewis, Anton Bakhtin, Yuxiang Wu, and Alexander Miller. 2019. Language Models as Knowledge Bases?. In Proceedings of the 2019 Conference on EMNLP and the 9th International Joint Conference on NLP. ACL, Hong Kong, China,...

    2019

  38. [46]

    Boyd-Graber, and Lijuan Wang

    Chenglei Si, Zhe Gan, Zhengyuan Yang, Shuohang Wang, Jianfeng Wang, Jor- dan L. Boyd-Graber, and Lijuan Wang. 2023. Prompting GPT-3 To Be Reliable. In The Eleventh International Conference on Learning Representations, ICLR 2023, Kigali, Rwanda, May 1-5, 2023 . OpenReview.net, Rwanda

    2023

  39. [47]

    Sneha Singhania, Simon Razniewski, and Gerhard Weikum. 2022. Predicting Document Coverage for Relation Extraction. TACL 10 (03 2022), 207–223

    2022

  40. [48]

    Sneha Singhania, Simon Razniewski, and Gerhard Weikum. 2024. Recall Them All: Retrieval-Augmented Language Models for Long Object List Extraction from Long Documents. ArXiv abs/2405.02732 (2024)

    2024 arXiv

  41. [49]

    Apoorv Saxena, Soumen Chakrabarti, and Partha Talukdar. 2021. Question Answering Over Temporal Knowledge Graphs. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on NLP (Volume 1: Long Papers...

    2021

  42. [50]

    Zhang, Makesh Narsimhan Sreedhar, and Oleksii Kuchaiev

    Zhilin Wang, Yi Dong, Olivier Delalleau, Jiaqi Zeng, Gerald Shen, Daniel Egert, Jimmy J. Zhang, Makesh Narsimhan Sreedhar, and Oleksii Kuchaiev. 2024. HelpSteer2: Open-source dataset for training top-performing reward models. arXiv:2406.08673

    2024 arXiv

  43. [51]

    Chi, Quoc V

    Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Brian Ichter, Fei Xia, Ed H. Chi, Quoc V. Le, and Denny Zhou. 2024. Chain-of-thought prompting elicits reasoning in large language models. InProceedings of the 36th International Conference on Neural Information Processin...

    2024

  44. [52]

    Siheng Xiong, Ali Payani, Ramana Kompella, and Faramarz Fekri. 2024. Large Language Models Can Learn Temporal Reasoning. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). ACL, Bangkok, Thailand, 10452–10470

    2024

  45. [53]

    Mengting Wan, Tara Safavi, Sujay Kumar Jauhar, Yujin Kim, Scott Counts, Jennifer Neville, Siddharth Suri, Chirag Shah, Ryen W White, Longqi Yang, Reid Andersen, Georg Buscher, Dhruv Joshi, and Nagu Rangan. 2024. TnT-LLM: Text Mining at Scale with Large Language Models. arXiv:2...

    2024 arXiv

  46. [57]

    It makes no sense for anyone to go as <e> Karl’s </e> cat more than <e> Doja Cat </e>,

    Wenda Xu, Danqing Wang, Liangming Pan, Zhenqiao Song, Markus Freitag, William Wang, and Lei Li. 2023. INSTRUCTSCORE: Towards Explainable Text Generation Evaluation with Automatic Feedback. In Proceedings of the 2023 Conference on Empirical Methods in NLP . ACL, Singapore, 5967...

    2023

  47. [58]

    rating": <numerical_rating>,

    {snippet} May 1, 2023: <e> Doja Cat </e> and <e> Jared Leto </e> deliver <e> Lagerfeld </e> -inspired looks. For this year's theme, the A-list guests were asked to wear outfits which paid homage to the German fashion designer … [k] {snippet} May 2 2023: … To be fair, <e> Leto <...

    2023

  48. [2012]

    In Proceedings of the 21st International Conference on WWW (Lyon, France)

    Active objects: actions for entity-centric search. In Proceedings of the 21st International Conference on WWW (Lyon, France). ACM, New York, 589–598

  49. [2020]

    InProceedings of the 37th ICML

    REALM: Retrieval-Augmented Language Model Pre-Training. InProceedings of the 37th ICML . JMLR.org, Online, Article 368, 10 pages

  50. [2022]

    TempoQR: Temporal Question Reasoning over Knowledge Graphs. InAAAI. The Thirty-Sixth AAAI Conference on Artificial Intelligence, Online, 5825–5833

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.