Pith. sign in

REVIEW 5 major objections 5 minor 30 references

CORG: Generating Answers from Complex, Interrelated Contexts

T0 review · 5 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read This paper claims that a graph-based, training-free context organizer raises disambiguated answer recall across eight models on question answering over complex, interrelated corpora.

desk verdict A useful taxonomy and broad experiments, but Algorithm 1's mirroring rule is logically wrong for mixed counterfactual/duplicated triples, so the mechanism as written can drop answers; still worthy of peer review. read the letter →

arxiv 2505.00023 v1 pith:ECI2PUBC submitted 2025-04-25 cs.CL cs.AI

Hyunji Lee , Franck Dernoncourt , Trung Bui , Seunghyun Yoon This is my paper · ORCID
classification cs.CLcs.AI
keywords multi-documentquestionansweringcontextrelationshipsdistractingcontextsambiguouscounterfactualduplicatedinference-timegroupingdisambiguation
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

Real-world corpora rarely present one clean answer: the same named entity can appear with different descriptors, with no descriptor at all, with contradictory facts, or with the same fact repeated verbatim. The paper classifies these as distracting, ambiguous, counterfactual, and duplicated relationships, and claims that no single prompt-level trick handles all four at once. It introduces CONTEXT ORGANIZER (CORG), an inference-time, training-free pipeline that builds a graph of pairwise context relations, reranks contexts into groups, and aggregates per-group, citation-backed answers. Across eight models of different sizes, CORG reports the highest disambiguated answer recall (Disambig-F1) among grouping methods on AmbigDocs+, ConflictQA+, and their single-factor bases, beating six baselines and roughly matching the far more expensive one-context-per-run method at lower cost. The practical claim is that how contexts are grouped for generation matters more than model scale, and that a cheap organization step can capture most of the benefit of full isolation.

What carries the argument

The load-bearing object is the graph constructor of Algorithm 1: it labels pairwise context relations with GPT-4, then propagates labels by assuming that when two contexts are counterfactual or duplicated, their relations to every third context are identical, so only missing edges need labeling in later iterations. The reranker consumes this graph and applies per-factor rules: for distracting pairs it removes the context lacking a descriptor; for duplicated pairs it keeps one randomly; and for counterfactual pairs it places each context in a separate group, distributing the remaining contexts across groups. The aggregator then runs each group through the model with a pluralized question and concatenates the outputs with citations. Together, the three components convert the four relationship types into a small number of independent inference runs, which is what produces the reported gains in answer recall and disambiguation.

What would settle it

Construct a hand-made triple in which context A is counterfactual with B and duplicated with C (same descriptor, A and C share an answer, A and B do not), run Algorithm 1, and check whether the B-C edge is labeled counterfactual or copied as duplicated. Since the correct relation is counterfactual, a duplicated label directly falsifies the propagation mechanism; a supplementary check is to manually label all pairwise relations on a random AmbigDocs+ sample and report the graph constructor's disagreement rate.

Watch

Extended reading notes

Core claim

The paper's central claim is that multi-document question answering over corpora with mixed context relationships is best served by separating counterfactual contexts into different inference groups, deduplicating repeated contexts, dropping ambiguous descriptor-less contexts when a distracting counterpart exists, and reformulating the question into a plural form so the model lists all answers. On AmbigDocs+ and ConflictQA+, CORG's grouping yields the best Disambig-F1—a model-based metric that checks whether the correct answer and its disambiguating descriptor both appear—among six baselines across all eight tested models, and it matches the much more expensive Separate method, which runs each context in its own forward pass. A supporting finding is that grouping by similarity (KMeans) is worse than random grouping, because similar-but-conflicting contexts blur entity descriptors; this locates the benefit in the graph's relationship labels rather than in clustering or plain input ordering.

Load-bearing premise

The load-bearing assumption is that once two contexts are marked counterfactual or duplicated, every other context relates to them in exactly the same way, so one context's relation labels can be copied onto the other without rechecking.

Editorial extensions

If this is right

  • On corpora that mix all four relationship types, an inference-time grouping step can beat both single-pass reading and similarity-based clustering without any model training.
  • Because each group's output carries citations, users can verify counterfactual or outdated claims from the source contexts instead of relying on the model to filter them.
  • The method transfers to newly released models directly, since its only adaptation is prompt-level graph labeling; the paper demonstrates consistent gains across eight models of different sizes.
  • Typical runs use roughly two inference groups per question rather than one group per context, so the answer-recall gains of separated processing are available at a small fraction of the cost.

Reading between the lines

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

  • The propagation rule in Algorithm 1 is not generally truth-preserving: if context A is counterfactual with B and duplicated with C, then B and C must be counterfactual, yet the copy rule labels B-C as duplicated. A direct check against human relation labels on AmbigDocs+ would reveal how often this error changes the groups and the final answer.
  • The paper reports no accuracy check for the GPT-4 relation labels feeding the graph. A controlled comparison using a cheaper labeler or human labels would separate the benefit of the grouping policy itself from the benefit of a labeler that happens to agree with the generator's biases.
  • The pluralization result points to a testable refinement: make question reformulation depend on the number of distinct descriptors in each group, rather than applying plural form to every multi-context group, which could reduce the over-generation the paper notes for single-context inputs.
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

5 major / 5 minor

Summary. The paper identifies four types of inter-context relationships in multi-document QA (distracting, ambiguous, counterfactual, duplicated), constructs extended datasets AmbigDocs+ and ConflictQA+ that mix all four factors, and analyzes simple per-factor solutions using Llama2-7B. It then proposes CORG, a training-free inference-time pipeline composed of a GPT-4-based graph constructor, a reranker that removes duplicates, drops descriptor-less contexts, and separates counterfactual contexts into groups, and an aggregator that pluralizes questions and concatenates cited answers. The paper reports D-F1 and related metrics across eight models and four datasets, claiming CORG consistently outperforms six baselines while matching the much more expensive Separate baseline, with the best performance-per-FLOPs trade-off on AmbigDocs+.

Significance. If the claims held, CORG would be a useful, practical contribution: an inference-time method that improves answer recall on mixed-relationship multi-document QA without training, with broad evaluation across model families and sizes. The paper includes human evaluation of the constructed datasets, ablations of the reranker/aggregator components, and a token-level efficiency analysis, which are valuable. However, the central mechanism has a documented logical inconsistency in the graph propagation rule, the evaluation design selects the per-factor solutions on the test corpora, the headline claim that CORG beats all six baselines is not supported by the reported tables, and the efficiency analysis excludes a per-query cost. These issues are load-bearing for the paper's main claims, though they may be addressable with additional validation and revised claims.

major comments (5)
  1. [Algorithm 1 (Graph Constructor), lines 8–11] The mirroring rule is logically inconsistent with the definitions in §2.1. When the newly processed context ci is counterfactual to the pivot c′ (same descriptor, different answer) and c′ is duplicated with cj (same descriptor, same answer), the edge (ci, cj) must be counterfactual under the paper's own taxonomy, but the algorithm copies the 'dup' edge, labeling ci and cj as duplicates. In the Appendix A.1 example, the main context (IIHF, Slovakia), the counterfactual context (IIHF, Canada), and the duplicated context (IIHF, Slovakia) form exactly this pattern; the reranker then randomly keeps one of the two 'duplicates' and loses an answer, contradicting the stated design goal of separating counterfactual contexts. Because no accuracy check of GPT-4 relation labels or of the resulting graph is reported, this flaw is not known to be corrected in the implementation and can directly undermine the claimed mechanism.
  2. [§3.2 and Tables 2–4; §5.3] The per-factor solutions (pluralizing the question, converting ambiguous contexts to distracting ones, separating counterfactual contexts, and deduplicating) are selected using Llama2-7B on AmbigDocs+ and ConflictQA+ instances, and the same corpora are then used for the final D-F1 evaluation in Tables 5, 6, and 10. The ablation in Table 11 is also on AmbigDocs+ with Llama2-7B. This means the design decisions are fit to the test distribution, and the reported gains over baselines may be optimistic. The paper should evaluate on a held-out development set or describe a model-selection procedure that does not use the test data.
  3. [Abstract; §5.3; Tables 5 and 10] The claim of 'consistently improves performance over six baselines across eight different models' is not supported by the reported numbers. In Table 5, Separate exceeds CORG in multiple rows: ConflictQA+ Llama2-13B (37.5 vs 30.8), Llama2-70B (40.1 vs 31.4), Llama3-8B (25.6 vs 22.8), Llama3-70B (35.2 vs 28.7), and ChatGPT (37.1 vs 35.9); in AmbigDocs+, Separate exceeds CORG for Llama2-70B (18.7 vs 17.9), Llama3-70B (22.4 vs 14.6), Mistral-7B (28.8 vs 27.5), and GPT-4o (32.1 vs 31.4). Similar patterns appear in Table 10 for ConflictQA. The accurate claim is that CORG outperforms grouping-based baselines and is comparable to Separate, not that it outperforms all six baselines. This should be corrected throughout the abstract and results sections.
  4. [§5.4, Figure 4, note 8] The efficiency analysis excludes graph construction from the FLOPs calculation, with the footnote stating that steps like graph construction are 'performed only once per dataset.' However, the graph constructor runs per question (Algorithm 1 takes the context list for a question), not once per dataset. Excluding this cost materially favors CORG in the performance/FLOPs comparison against baselines that do not perform relation extraction. The paper should either include the graph-construction cost or explicitly state that the comparison is for generation FLOPs only, then discuss the implication for the practical efficiency claim.
  5. [Appendix C.1 and Figure 9] The graph constructor uses an input format that adds a 'None' relationship (different descriptor with one being Null and different answer), which is not part of the four-type taxonomy in §2.1, and the prompt example in Figure 9 labels the pair 'Context2 - None' even though Context2 (Slovakia) and Context1 (Canada) appear to satisfy the counterfactual definition (same entity, different answers). This ambiguity in the labeling schema, combined with the absence of any accuracy or inter-annotator agreement measure for the GPT-4 relation labels, makes it impossible to verify that the constructed graph matches the paper's definitions. Please report the accuracy of relation labeling on a sample and reconcile the 'None' category with §2.1.
minor comments (5)
  1. [§5.2] The text says 'We evaluate five baselines' but then lists six (Base, Retrieve, Summarize, Random, KMeans, Separate); §5.3 correctly says 'six baselines.' Please harmonize the count.
  2. [§3.1 heading] The heading 'Affect of each factor' should be 'Effect of each factor.'
  3. [§4, Reranker paragraph] The reranker description says 'for contexts in a distracting relationship, when a context with a descriptor is available, we remove the one without it,' but a pair in which one context lacks a descriptor is ambiguous, not distracting; this appears to conflict with §3.2's 'Change to Distracting' solution and should be clarified.
  4. [Table 11] Variant (2) is labeled 'without converting distracting to ambiguous,' which is the reverse of the §3.2 solution 'Change to Distracting'; please align the terminology with the prose.
  5. [Figure numbering] Section 3.1 refers to 'Figure 3' for performance trends, while Figure 3 is used for the CORG system overview; the cross-references and figure numbers should be fixed.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: CORG's improvements are empirical benchmark measurements; the flagged Algorithm 1 propagation flaw is a correctness issue, not a circular reduction.

full rationale

I examined the claimed derivation chain: the four relationship definitions in Section 2.1, the per-factor analyses in Section 3, the CORG algorithm in Section 4, and the D-F1 results in Section 5. The central claim is an empirical comparison against six baselines on four datasets; it is not derived by construction from the relationship taxonomy or from Algorithm 1. No equation equates a fitted value with a reported prediction. The per-factor rule choices (pluralization, converting ambiguous to distracting, separating counterfactuals, deduplication) were developed using the same corpus family on which final numbers are reported, which is a data-snooping and selection concern that could inflate results, but the final measurements are not statistically forced by the rules themselves. The self-citations to AmbigDocs (Lee et al., 2024c) are dataset and metric references, not load-bearing circular support. The logical flaw in Algorithm 1's mirroring of counterfactual edges (e.g., copying a 'duplicated' edge from a pivot to a counterfactual context, although the definitions imply 'counterfactual') is a genuine correctness bug that could undermine the mechanism, but it is not a circularity: the reported gains are not equivalent to the input labels by construction. I therefore find no significant circularity.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central claim rests on a taxonomy of context relationships, a set of per-factor heuristics selected on the same benchmarks used for evaluation, and a GPT-4-based graph constructor whose output is not validated. No numeric free parameters are fitted, but the design choices (pluralization, descriptor replacement, separation, deduplication) are tuned on Llama2-7B using the test distribution.

assumptions (5)
  • domain assumption Contexts can be decomposed into surface name, descriptor, and answer, and every inter-context relationship falls into exactly one of four categories: ambiguous, distracting, counterfactual, duplicated.
    Section 2.1 defines relationships this way; the whole grouping strategy depends on this taxonomy covering real-world cases.
  • ad hoc to paper Replacing an ambiguous context (missing descriptor) with a descriptor-bearing version preserves the answer and improves performance.
    Table 3, 'Change to Distracting', is used in the reranker; the paper acknowledges not all ambiguous contexts can be replaced (Section 4 footnote), yet the method removes the one without descriptor when possible.
  • ad hoc to paper Separating counterfactual contexts into distinct inference runs improves answer recall.
    Established only on Llama2-7B in Table 4; applied to all models in CORG.
  • domain assumption GPT-4 reliably classifies context relationships for graph construction.
    No accuracy evaluation of the graph constructor is reported; the prompts in Figure 9 are the only specification.
  • ad hoc to paper If context A is counterfactual or duplicated with B, then B's relationships to all other contexts mirror A's relationships to them.
    Algorithm 1 lines 8-11; this logical rule is not proven and is false in the mixed case described in the weakest_assumption.

how reviews work

0 comments
Cite this review

Pith. "Pith review of CORG: Generating Answers from Complex, Interrelated Contexts." pith.science (2026). https://pith.science/paper/ECI2PUBC

@misc{pith2026250500023,
  author       = {Pith},
  title        = {Pith review of: CORG: Generating Answers from Complex, Interrelated Contexts},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ECI2PUBC}},
  note         = {Machine review of arXiv:2505.00023}
}
read the original abstract

In a real-world corpus, knowledge frequently recurs across documents but often contains inconsistencies due to ambiguous naming, outdated information, or errors, leading to complex interrelationships between contexts. Previous research has shown that language models struggle with these complexities, typically focusing on single factors in isolation. We classify these relationships into four types: distracting, ambiguous, counterfactual, and duplicated. Our analysis reveals that no single approach effectively addresses all these interrelationships simultaneously. Therefore, we introduce Context Organizer (CORG), a framework that organizes multiple contexts into independently processed groups. This design allows the model to efficiently find all relevant answers while ensuring disambiguation. CORG consists of three key components: a graph constructor, a reranker, and an aggregator. Our results demonstrate that CORG balances performance and efficiency effectively, outperforming existing grouping methods and achieving comparable results to more computationally intensive, single-context approaches.

Figures

Figures reproduced from arXiv: 2505.00023 by the authors.

Figure 1
Figure 1. In real-world corpora, contexts often exhibit com￾plex interrelationships, which we classify into four categories: distracting, counterfactual, duplicated, and ambiguous. “The Simpsons” is the surface name, and “Season 2” is the descriptor that specifies the entity. Based on these attributes, we classify contexts as distract￾ing (same surface name, different descriptors), am￾biguous (same surface name, only one with… view at source ↗
Figure 2
Figure 2. Answer recall as the number of contexts increases for each factor. For ambiguous cases, note that since ambiguity can only exist between two contexts, sets with three or more contexts (x > 2) also include distracting relationships. 3 Analyzing Solution for each Factor In this section, we analyze how adding the con￾text of each factor to the input affects model per￾formance (Section 3.1) and investigate a simple solu… view at source ↗
Figure 3
Figure 3. Overview of CONTEXT ORGANIZER (CORG), composed of three components: graph constructor, reranker, and aggregator. Given a corpus with multiple relevant contexts and a question, the graph constructor and the reranker organize the corpus based on the question, and the aggregator generates a response containing all possible answers with references. Algorithm 1 Graph Constructor 1: Input: List of contexts C = [c0, c1, . … view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Trade-off between efficiency and performance for Llama2-7B in AmbigDocs+. Light color indicates better balance between the two; CORG shows the best followed by Summarize. lack of descriptor emphasis limits effective entity disambiguation, whereas methods processing mul…
Figure 5
Figure 5. Figure 5: Instruction to Human Evaluators A.4 Details of corpus construction To evaluate the LLM’s ability with corpora rep￾resenting real-world scenarios, we generate addi￾tional contexts based on existing datasets consist￾ing of (question, answer, contexts) pairs. We use Ambig…
Figure 6
Figure 6. Figure 6: Prompt to Generate Context cated contexts to AmbigDocs, and add distracting, ambiguous, and duplicated contexts to ConflictQA, which we name as AmbigDocs+ and ConflictQA+, respectively. In detail, AmbigDocs contains pairs of question q and a list of contexts relevant t…
Figure 7
Figure 7. Figure 7: Prompt to Generate Sub-Questions for ConflictQA [PITH_FULL_IMAGE:figures/full_fig_p014_7.png]
Figure 8
Figure 8. Figure 8: Statistics showing the number of questions (y-axis) against the number of contexts per question (x-axis). Most cases have eight contexts for each question. flictQA and ConflictQA+ datasets include 1,000 evaluation samples. B Analyzing Solution for each Factor B.1 Ablat…
Figure 9
Figure 9. Figure 9: Input format to graph constructor [PITH_FULL_IMAGE:figures/full_fig_p015_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 10 canonical work pages

  1. [1]

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. 2023. Gpt-4 technical report. arXiv preprint arXiv:2303.08774

    arXiv 2023

  2. [2]

    Zhang, and Eunsol Choi

    Hung-Ting Chen, Michael J.Q. Zhang, and Eunsol Choi. 2022. https://api.semanticscholar.org/CorpusID:253107178 Rich knowledge sources bring complex knowledge conflicts: Recalibrating models to reflect conflicting evidence . In Conference on Empirical Methods in Natural Language Processing

    work page 2022

  3. [3]

    Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. https://api.semanticscholar.org/CorpusID:52967399 Bert: Pre-training of deep bidirectional transformers for language understanding . In North American Chapter of the Association for Computational Linguistics

    2019

  4. [4]

    Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Amy Yang, Angela Fan, et al. 2024. The llama 3 herd of models. arXiv preprint arXiv:2407.21783

    arXiv 2024

  5. [5]

    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 preprint arXiv:2404.16130

    arXiv 2024

  6. [6]

    Gizem Gezici, Aldo Lipani, Yucel Saygin, and Emine Yilmaz. 2021. Evaluation metrics for measuring bias in search engine results. Information Retrieval Journal, 24:85--113

    work page 2021

  7. [7]

    Shibo Hao, Tianyang Liu, Zhen Wang, and Zhiting Hu. 2024. Toolkengpt: Augmenting frozen language models with massive tools via tool embeddings. Advances in neural information processing systems, 36

    2024

  8. [8]

    Cheng-Mao Hsu, Cheng te Li, Diego S \'a ez-Trumper, and Yi-Zhan Hsu. 2021. https://api.semanticscholar.org/CorpusID:244130115 Wikicontradiction: Detecting self-contradiction articles on wikipedia . 2021 IEEE International Conference on Big Data (Big Data), pages 427--436

    work page 2021

Show all 30 references
  1. [9]

    Albert Q Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lucile Saulnier, et al. 2023. Mistral 7b. arXiv preprint arXiv:2310.06825

    2023 arXiv

  2. [10]

    Zhuoran Jin, Pengfei Cao, Yubo Chen, Kang Liu, Xiaojian Jiang, Jiexin Xu, Qiuxia Li, and Jun Zhao. 2024. https://api.semanticscholar.org/CorpusID:267782658 Tug-of-war between knowledge: Exploring and resolving knowledge conflicts in retrieval-augmented language models . ArXiv,...

    2024 arXiv

  3. [11]

    Hyunji Lee, Sejune Joo, Chaeeun Kim, Joel Jang, Doyoung Kim, Kyoung-Woon On, and Minjoon Seo. 2023. How well do large language models truly ground? NACCL 2024

    2023

  4. [12]

    Hyunji Lee, Luca Soldaini, Arman Cohan, Minjoon Seo, and Kyle Lo. 2024 a . Routerretriever: Exploring the benefits of routing over multiple expert embedding models. arXiv preprint arXiv:2409.02685

    2024 arXiv

  5. [13]

    Jinhyuk Lee, Anthony Chen, Zhuyun Dai, Dheeru Dua, Devendra Singh Sachan, Michael Boratko, Yi Luan, S \'e bastien MR Arnold, Vincent Perot, Siddharth Dalmia, et al. 2024 b . Can long-context language models subsume retrieval, rag, sql, and more? arXiv preprint arXiv:2406.13121

    2024 arXiv

  6. [14]

    Yoonsang Lee, Xi Ye, and Eunsol Choi. 2024 c . Ambigdocs: Reasoning across documents on different entities under the same name. COLM 2024

    2024

  7. [15]

    Longpre, Kartik Perisetla, Anthony Chen, Nikhil Ramesh, Chris DuBois, and Sameer Singh

    S. Longpre, Kartik Perisetla, Anthony Chen, Nikhil Ramesh, Chris DuBois, and Sameer Singh. 2021. https://api.semanticscholar.org/CorpusID:237491581 Entity-based knowledge conflicts in question answering . ArXiv, abs/2109.05052

    2021 arXiv

  8. [16]

    Sewon Min, Kenton Lee, Ming-Wei Chang, Kristina Toutanova, and Hannaneh Hajishirzi. 2021. Joint passage ranking for diverse multi-answer retrieval. arXiv preprint arXiv:2104.08445

    2021 arXiv

  9. [17]

    Sewon Min, Julian Michael, Hannaneh Hajishirzi, and Luke Zettlemoyer. 2020. A mbig QA : Answering ambiguous open-domain questions. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)

    2020

  10. [18]

    Bhaskarjit Sarmah, Benika Hall, Rohan Rao, Sunil Patel, Stefano Pasquali, and Dhagash Mehta. 2024. https://api.semanticscholar.org/CorpusID:271843286 Hybridrag: Integrating knowledge graphs and vector retrieval augmented generation for efficient information extraction . In Int...

    2024

  11. [19]

    Timo Schick, Jane Dwivedi-Yu, Roberto Dess \` , Roberta Raileanu, Maria Lomeli, Eric Hambro, Luke Zettlemoyer, Nicola Cancedda, and Thomas Scialom. 2024. Toolformer: Language models can teach themselves to use tools. Advances in Neural Information Processing Systems, 36

    2024

  12. [20]

    Ivan Stelmakh, Yi Luan, Bhuwan Dhingra, and Ming-Wei Chang. 2022. Asqa: Factoid questions meet long-form answers. arXiv preprint arXiv:2204.06092

    2022 arXiv

  13. [21]

    Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, et al. 2023. Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:2307.09288

    2023 arXiv

  14. [22]

    Xiangchen Wu, Xi Niu, and Ruhani Rahman. 2022. https://api.semanticscholar.org/CorpusID:250340379 Topological analysis of contradictions in text . Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval

    2022

  15. [23]

    Jian Xie, Kai Zhang, Jiangjie Chen, Renze Lou, and Yu Su. 2023. https://api.semanticscholar.org/CorpusID:263610324 Adaptive chameleon or stubborn sloth: Revealing the behavior of large language models in knowledge conflicts

    2023

  16. [24]

    Fangyuan Xu, Weijia Shi, and Eunsol Choi. 2023. Recomp: Improving retrieval-augmented lms with compression and selective augmentation. arXiv preprint arXiv:2310.04408

    2023 arXiv

  17. [25]

    Rongwu Xu, Zehan Qi, Cunxiang Wang, Hongru Wang, Yue Zhang, and Wei Xu. 2024. Knowledge conflicts for llms: A survey. arXiv preprint arXiv:2403.08319

    2024 arXiv

  18. [26]

    Shunyu Yao, Jeffrey Zhao, Dian Yu, Nan Du, Izhak Shafran, Karthik Narasimhan, and Yuan Cao. 2022. React: Synergizing reasoning and acting in language models. arXiv preprint arXiv:2210.03629

    2022 arXiv

  19. [27]

    Zhang and Eunsol Choi

    Michael J.Q. Zhang and Eunsol Choi. 2021. https://api.semanticscholar.org/CorpusID:237491751 Situatedqa: Incorporating extra-linguistic contexts into qa . ArXiv, abs/2109.06157

    2021 arXiv

  20. [28]

    Wenxuan Zhou, Sheng Zhang, Hoifung Poon, and Muhao Chen. 2023. https://api.semanticscholar.org/CorpusID:257632259 Context-faithful prompting for large language models . In Conference on Empirical Methods in Natural Language Processing

    2023

  21. [29]

    URL: " 'urlintro :=

    ENTRY address author booktitle chapter edition editor howpublished institution journal key month note number organization pages publisher school series title type volume year eprint doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before...

  22. [30]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.