Pith. sign in

REVIEW 3 major objections 5 minor 26 references

Enhancing Entertainment Translation for Indian Languages using Adaptive Context, Style and LLMs

T0 review · 3 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read This paper claims that feeding a large language model a per-scene estimate of plot context and style, rather than a fixed creative prompt, yields consistently better entertainment translations across eight LLMs, as measured by…

desk verdict A plausible prompt-engineering recipe for Indian-language entertainment translation, but the evaluation is too fragile and the manuscript too unfinished to accept. read the letter →

arxiv 2412.20440 v1 pith:KLWZR7YB submitted 2024-12-29 cs.CL

Pratik Rakesh Singh , Mohammadi Zaki , Pankaj Wasnik This is my paper · ORCID
classification cs.CL
keywords entertainmenttranslationneuralmachinelargelanguagemodelscontext-awarepromptingstyleadaptationretrieval-augmentedgenerationIndianlanguagessubtitle
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

This paper tries to show that movie and series translation into Indian languages improves substantially when the prompt given to a large language model is rebuilt for each scene from two estimates: the current plot context and the scene's style. It introduces CASAT, a pipeline that splits a script into tonal sessions, retrieves plot summaries relevant to the line being translated from a vector store, and extracts style markers such as word-class frequencies, modal words, idioms, intent, and emotion. On web-scraped subtitles for English-to-Hindi, English-to-Bengali, and English-to-Telugu, CASAT-augmented LLMs outperform their own baseline prompts and traditional MT systems on reference-free COMET and on a GPT-4o win-ratio, across eight LLMs from 7B to 70B parameters. The significance is that the method requires no timestamps, speaker IDs, or retraining, so it could be a drop-in wrapper for existing localization pipelines.

What carries the argument

The central object is the time-varying prompt $p_t = h(\tilde{c}_t, \tilde{s}_x)$ that adapts to the current session's estimated context $\tilde{c}_t$ and style $\tilde{s}_x$. The context estimate comes from a retrieval-augmented generation module: an offline LLM creates plot summaries of K-session windows, chunks them into a vector database, and a retriever-reranker selects the N most relevant chunks for the source sentence. The style estimate comes from a Domain Adaptation Module that pools dialogue-level statistics (content/function word distributions via an Indic POS tagger, frequent syllabic words, modal words, idioms) with session-level signals (punctuation-derived intent and LLM-extracted emotion). The session segmentation algorithm (Alg 1) is what makes these estimates time-dependent, by grouping sentences into nearly constant-tone sessions; without it the prompt collapses to a static style hint.

What would settle it

A blind human preference study on the same test sets would settle the claim: if bilingual viewers or professional subtitle translators do not choose CASAT translations over baseline LLM outputs at a rate consistent with the reported win-ratio, the measured gains would not reflect audience-relevant quality.

Watch

Extended reading notes

Core claim

The central claim is that context and style are not static properties of entertainment content but vary from scene to scene, so a translation prompt conditioned on static instructions misses what a line needs. CASAT operationalizes this by generating context and style estimates per session. A session classifier groups sentences into serious, casual, or neutral segments using cosine similarity over sentence embeddings, with windowing to keep sessions coherent. An offline LLM writes plot summaries for K consecutive sessions, which are chunked and indexed in a vector database; at translation time a retriever and reranker select the N most relevant chunks as context. In parallel, a Domain Adaptation Module computes style descriptors: part-of-speech distributions for content versus function words, frequent monosyllabic and polysyllabic words, modal particles and idioms, punctuation-derived intent, and an LLM-generated emotion label. These estimates are concatenated into the prompt that guides the LLM's translation. The experiments report that this adaptive prompt beats a fixed creative prompt, that plot context alone improves relevance while style alone improves literal fidelity, and that combining the two gives the best COMET and win-ratio results across all tested LLMs.

Load-bearing premise

The load-bearing premise is that reference-free COMET and a GPT-4o preference judge measure what audiences actually value in entertainment translation, since no human evaluation is included in the study.

Editorial extensions

If this is right

  • Entertainment translation into Indian languages can be improved without retraining models: wrapping off-the-shelf LLMs with session-adaptive prompts raises COMET scores and win-ratios over both static creative prompts and traditional MT systems like IndicTrans2 and NLLB.
  • The method generalizes across model families and sizes because it only changes the prompt; the paper reports gains across eight LLMs from 7B to 70B and three language pairs (English to Hindi, Bengali, and Telugu).
  • Context and style contribute complementary gains: plot context improves relevance as measured by COMET and win-ratio, style improves literal fidelity as measured by BLEU, and the full CASAT prompt yields the best combined scores.
  • Retrieving plot chunks from a vector store outperforms passing the surrounding sentences as context, indicating that non-adjacent scenes often carry the decisive information for a line, and that the retrieval window should not be limited to immediate neighbors.
  • Larger LLMs exploit the adaptive prompt better, so the quality gap between base and CASAT-augmented translation grows with model capacity.

Reading between the lines

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

  • If the win-ratio reflects a genuine audience preference, the same session-aware prompt construction could be applied to other creative-generation tasks, such as dubbing scripts segmentation or dialogue rewriting, where style descriptors are already estimated.
  • The paper's style descriptors were designed for Indian languages; a testable extension would be running CASAT on other language families to see whether the same descriptors (syllable patterns, modal words) transfer or need to be redesigned.
  • The authors note that the offline plot-design component is a target for removal; if the whole pipeline could run online, live streaming localization would become the natural deployment scenario.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes CASAT (Context And Style Aware Translation), a pipeline for translating entertainment dialogue without timestamps or speaker IDs. CASAT segments a movie or series into 'sessions' by tonal genre, uses an LLM to write plot summaries for nearby sessions, retrieves relevant plot chunks via a reranking RAG module, and extracts style descriptors (POS-based content/function words, syllabic word frequencies, modal words/idioms, intent, emotion) through a Domain Adaptation Module. These context and style signals are concatenated into a prompt for an LLM translator. The method is evaluated on web-scraped parallel subtitle data for English-to-Hindi, English-to-Bengali, and English-to-Telugu, across seven LLM baselines, with BLEU, reference-free COMET (wmt22-cometkiwi-da), and a GPT-4o-based win-ratio. The authors report COMET gains of roughly 0.01--0.05 and win-ratios mostly in the 0.6--0.85 range, and conclude that context and style awareness consistently improve LLM-based entertainment translation.

Significance. If the empirical claims were reliable, CASAT would be a practical, language-agnostic recipe for improving LLM translation in the under-served entertainment domain, with the useful property of not depending on timestamps or speaker identification. Strengths of the submission include a clear problem formulation, explicit pseudocode for the algorithm, a hyperparameter table (Table 6), and ablation experiments (Table 2) that attempt to attribute gains to context versus style. However, the central claim is empirical, and the evidence as presented is not yet convincing: the automatic metrics are used in a regime the paper itself acknowledges is problematic for creative translation, the win-ratio judge is not validated, and no uncertainty quantification is provided. The significance would be much higher with a human evaluation or a validated creative-translation metric.

major comments (3)
  1. [Experiments, Evaluation Metrics; Table 1] The abstract's claim of 'significant improvement' and 'consistently outperforms' is not supported by any significance testing, confidence intervals, or error bars. The COMET deltas in Table 1 are mostly 0.01 to 0.05, which is plausibly within metric noise for 800 samples per language direction. The win-ratio uses GPT-4o as a judge, but the paper does not specify the judge prompt, tie handling, order randomization, or any human correlation; the cited GPTScore work does not establish that GPT-4o is a valid judge of dubbing-style preference. Without these controls, the central empirical claim is not established.
  2. [Appendix, 'Possible Reasons for Low Automatic metric scores'; Table 1] The manuscript itself concedes that entertainment translations in this dataset 'are rarely literal or word-for-word' and 'often result in lower BLEU scores,' yet BLEU and COMET are used as headline evidence in Tables 1 and 2. This is a metric-validity problem: the one-to-many mapping of creative translation means a low BLEU or moderate COMET may not indicate poor quality. Table 1 provides a direct internal contradiction: for Mistral 7B En-Tel, CASAT decreases BLEU (0.10 to 0.07) and COMET (0.42 to 0.41) while the GPT-4o win-ratio is 0.8, so at least one of the metrics is not tracking the intended construct. A human evaluation, or a validated metric for creative subtitles, is needed to support the headline claims.
  3. [Appendix] Several parts of the appendix referenced in the main text are missing or incomplete. The main text refers to 'the exact prompt used for baseline LLMs' in the Appendix, but no such prompt appears. The subsection 'Choice of hyper-parameters' is empty. Table 7 has a placeholder caption for a 13-column table with no contents. These omissions prevent the reader from verifying the experimental setup and reproducing the comparisons, which is load-bearing for the central empirical claim.
minor comments (5)
  1. [Abstract and Experiments] The abstract claims the method is 'language and LLM-agnostic,' but the experiments cover only English-to-Indian-language directions (En-Hi, En-Ben, En-Tel). This overstates the empirical scope; the claim should be softened or supported by additional language pairs.
  2. [Methodology, 'Adaptive Session Classification and Segmentation'] There is a duplicated phrase: 'classifies each session into one of the three three primary tonal categories.' Please remove the extra 'three'.
  3. [Table 3] The header of Table 3 reads 'Wini-ratio' instead of 'Win-ratio.'
  4. [Appendix, 'Additional Ablation Experiments'] The sentence 'As detailed in Section ,' contains an empty cross-reference; it should point to the specific section or be reworded.
  5. [Figure 4] The example text inside Figure 4 appears to repeat the same clause ('The conversation's have many long sentences, rich connotations, specific narratives, detailed reasoning, and full of emotions') twice in different boxes; verify that this is not a copy-paste artifact.

Circularity Check

0 steps flagged · score 0.0 of 10

No constructional circularity: CASAT's COMET and win-ratio claims are empirical measurements, not identities forced by the method's definitions; K-selection and LLM-judge concerns are validity risks, not circular steps.

full rationale

I walked the derivation chain from the abstract's claims of 'significant improvement in the COMET scores' and 'consistently outperforms baseline LLMs in terms of win-ratio.' CASAT (Algorithm 2) takes source sentences, segments them via the k-NN genre classifier (Algorithm 1), retrieves ranked plot chunks through the retriever/reranker, extracts style statistics through the Domain Adaptation Module, concatenates these into prompt pt, and calls an LLM to produce yx. No equation in the paper defines yx as the evaluation metric, and no fitted parameter is renamed as a prediction: the win-ratio Δ is an explicitly defined empirical count (Section Evaluation Metrics), while COMET (wmt22-cometkiwi-da) is an external reference-free model whose scores are not used to construct the prompts. The selection of K by inspecting COMET on the same En-Hi data (Table 4) is a test-set selection risk, and the appendix's admission that creative translations 'are rarely literal or word-for-word' and 'often result in lower BLEU scores' is an evaluation-validity concern, but neither step reduces the central claim to its own inputs. The reference list contains no self-citations by the present authors; prior work is cited for components (e.g., CAT-LLM, RAG, reranking), not to license the headline conclusion. I therefore find no significant constructional circularity.

Assumptions & free parameters 9 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the validity of the segmentation, the accuracy of LLM-generated plot/emotion descriptors, the adequacy of word-frequency statistics as style representations, and the reliability of reference-free COMET and GPT-4o as quality metrics. These are domain assumptions, not standard mathematical axioms. Several hyper-parameters (K, M, N, alpha, beta, temperatures, chunking) are chosen by hand or tuned on the test data; none are derived from theory.

free parameters (9)
  • K (number of sessions used for context and style) = 2
    Selected because K=2 and K=3 gave the best COMET scores in Table 4; the paper uses K=2 for all experiments.
  • M (number of retrieved context chunks) = 5
    Chosen in the hyper-parameter table with no sensitivity analysis.
  • N (number of reranked context chunks) = 2
    Chosen in the hyper-parameter table with no sensitivity analysis.
  • alpha (minimum sentences before a genre switch is accepted) = 5
    Set in the hyper-parameter table; controls session segmentation in Alg.1.
  • beta (maximum sentences in a session) = 10
    Set in the hyper-parameter table; controls session segmentation in Alg.1.
  • chunk size / overlap for vector database = 356 / 64
    Set in the hyper-parameter table.
  • temperature for plot design LLM = 0.5
    Set in the hyper-parameter table.
  • temperature for emotion and translation LLM = 0.2
    Set in the hyper-parameter table.
  • intent punctuation thresholds = unspecified
    The DAM module defines sentence intent based on punctuation thresholds, but the actual threshold values are not provided.
assumptions (4)
  • domain assumption Movie dialogues can be partitioned into sessions of stable tone (Serious, Casual, Neutral) using k-NN on sentence embeddings with synthetic exemplars.
    Invoked in 'Adaptive Session Classification and Segmentation' and Alg.1; if segmentation is wrong, context/style estimates are unreliable.
  • domain assumption LLM-generated plot summaries and emotion labels for each session are accurate enough to guide translation.
    Used in the Context Retrieval module and Session Level Module; no validation of these intermediate outputs is provided.
  • ad hoc to paper The style of a dialogue can be captured by frequency statistics of POS-tagged content/function words, syllabic word counts, modal words, and idioms from past translations.
    The Domain Adaptation Module relies on this representational assumption; no evidence that these statistics correlate with perceived style is given.
  • domain assumption Reference-free COMET (wmt22-cometkiwi-da) and GPT-4o win-ratio are valid proxies for translation quality in creative entertainment content.
    Used in Evaluation Metrics; no human evaluation or correlation study is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Enhancing Entertainment Translation for Indian Languages using Adaptive Context, Style and LLMs." pith.science (2026). https://pith.science/paper/KLWZR7YB

@misc{pith2026241220440,
  author       = {Pith},
  title        = {Pith review of: Enhancing Entertainment Translation for Indian Languages using Adaptive Context, Style and LLMs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KLWZR7YB}},
  note         = {Machine review of arXiv:2412.20440}
}
read the original abstract

We address the challenging task of neural machine translation (NMT) in the entertainment domain, where the objective is to automatically translate a given dialogue from a source language content to a target language. This task has various applications, particularly in automatic dubbing, subtitling, and other content localization tasks, enabling source content to reach a wider audience. Traditional NMT systems typically translate individual sentences in isolation, without facilitating knowledge transfer of crucial elements such as the context and style from previously encountered sentences. In this work, we emphasize the significance of these fundamental aspects in producing pertinent and captivating translations. We demonstrate their significance through several examples and propose a novel framework for entertainment translation, which, to our knowledge, is the first of its kind. Furthermore, we introduce an algorithm to estimate the context and style of the current session and use these estimations to generate a prompt that guides a Large Language Model (LLM) to generate high-quality translations. Our method is both language and LLM-agnostic, making it a general-purpose tool. We demonstrate the effectiveness of our algorithm through various numerical studies and observe significant improvement in the COMET scores over various state-of-the-art LLMs. Moreover, our proposed method consistently outperforms baseline LLMs in terms of win-ratio.

Figures

Figures reproduced from arXiv: 2412.20440 by the authors.

Figure 1
Figure 1. Examples of common mistakes made by NMT sys [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. A high-level overview of our proposed methodology. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. A block diagram of the Context retriever block. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Domain Adaptation Module Style extraction-Domain Adaptation Module: By using the above pipeline for context information extraction, we [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 5
Figure 5. Figure 5: A comparative analysis of the effect of various prompts on the translated text. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Clusters obtained using k-NN for genre segmentation. [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]
Figure 7
Figure 7. Figure 7: Samples of Source and Reference of the dataset [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]
Figure 8
Figure 8. Figure 8: Explicit prompts used for the plot designer, session emotion extractor, and the final translation generator LLMs. [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 9 canonical work pages

  1. [5]

    arXiv:2407.06204

    A Survey on Mixture of Experts. arXiv:2407.06204. Carnap, R

    arXiv

  2. [7]

    arXiv:2302.04166

    GPTScore: Evaluate as You Desire. arXiv:2302.04166. Gaido, M.; Papi, S.; Negri, M.; Cettolo, M.; and Bentivogli, L

    arXiv

  3. [8]

    SBAAM! Eliminating Transcript Dependency in Automatic Subtitling

    SBAAM! Eliminating Transcript Dependency in Automatic Subtitling. arXiv:2405.10741. Gala, J.; Chitale, P. A.; Raghavan, A. K.; Gumma, V .; Dod- dapaneni, S.; M, A. K.; Nawale, J. A.; Sujatha, A.; Pudup- pully, R.; Raghavan, V .; Kumar, P.; Khapra, M. M.; Dabre, R.; and Kunchukuttan, A

    work page Pith review arXiv

  4. [9]

    In Bouamor, H.; Pino, J.; and Bali, K., eds., Findings of the Association for Compu- tational Linguistics: EMNLP 2023 , 1078–1089

    An Adaptive Prompt Generation Framework for Task-oriented Dialogue System. In Bouamor, H.; Pino, J.; and Bali, K., eds., Findings of the Association for Compu- tational Linguistics: EMNLP 2023 , 1078–1089. Singapore: Association for Computational Linguistics. Gemma-Team

    work page 2023

  5. [10]

    arXiv:2408.00118

    Gemma 2: Improving Open Language Models at a Practical Size. arXiv:2408.00118. Glass, M.; Rossiello, G.; Chowdhury, M. F. M.; Naik, A. R.; Cai, P.; and Gliozzo, A

    arXiv

  6. [13]

    arXiv:2310.06825

    Mistral 7B. arXiv:2310.06825. Karakanta, A.; Bentivogli, L.; Cettolo, M.; Negri, M.; and Turchi, M

    arXiv

  7. [14]

    arXiv:2309.06085

    BHASA: A Holis- tic Southeast Asian Linguistic and Cultural Evaluation Suite for Large Language Models. arXiv:2309.06085. Lewis, P.; Perez, E.; Piktus, A.; Petroni, F.; Karpukhin, V .; Goyal, N.; Küttler, H.; Lewis, M.; tau Yih, W.; Rock- täschel, T.; Riedel, S.; and Kiela, D

    arXiv

  8. [16]

    In Màrquez, L.; Callison-Burch, C.; and Su, J., eds., Pro- ceedings of the 2015 Conference on Empirical Methods in Natural Language Processing , 1412–1421

    Effective Approaches to Attention-based Neural Machine Translation. In Màrquez, L.; Callison-Burch, C.; and Su, J., eds., Pro- ceedings of the 2015 Conference on Empirical Methods in Natural Language Processing , 1412–1421. Lisbon, Portu- gal: Association for Computational Linguistics. Lyu, C.; Du, Z.; Xu, J.; Duan, Y .; Wu, M.; Lynn, T.; Aji, A. F.; Wong...

    work page 2015

Show all 26 references
  1. [17]

    A Paradigm Shift: The Future of Machine Translation Lies with Large Language Models. In Calzolari, N.; Kan, M.-Y .; Hoste, V .; Lenci, A.; Sakti, S.; and Xue, N., eds., Proceedings of the 2024 Joint International Conference on Compu- tational Linguistics, Language Resources an...

    2024

  2. [18]

    arXiv:2407.21783

    The Llama 3 Herd of Models. arXiv:2407.21783. Post, M

    arXiv

  3. [19]

    arXiv:2209.06243

    CometKiwi: IST-Unbabel 2022 Submission for the Quality Estimation Shared Task. arXiv:2209.06243. Salinas, A.; and Morstatter, F

    2022 arXiv

  4. [20]

    arXiv:2401.03729

    The Butterfly Effect of Altering Prompts: How Small Changes and Jailbreaks Affect Large Language Model Performance. arXiv:2401.03729. Tao, Z.; Xi, D.; Li, Z.; Tang, L.; and Xu, W

    arXiv

  5. [21]

    arXiv:2401.05707

    CAT-LLM: Prompting Large Language Models with Text Style Definition for Chinese Article-style Transfer. arXiv:2401.05707. Team, N.; Costa-jussà, M. R.; Cross, J.; Çelebi, O.; El- bayad, M.; Heafield, K.; Heffernan, K.; Kalbassi, E.; Lam, J.; Licht, D.; Maillard, J.; Sun, A.; W...

    arXiv

  6. [22]

    arXiv:2207.04672

    No Language Left Behind: Scaling Human-Centered Machine Translation. arXiv:2207.04672. Vaswani, A.; Shazeer, N.; Parmar, N.; Uszkoreit, J.; Jones, L.; Gomez, A. N.; Kaiser, L. u.; and Polosukhin, I

    arXiv

  7. [25]

    arXiv:2305.14328

    Benchmark- ing LLM-based Machine Translation on Cultural Aware- ness. arXiv:2305.14328. Zhang, B.; Haddow, B.; and Birch, A

    arXiv

  8. [26]

    arXiv:2309.01219

    Siren’s Song in the AI Ocean: A Survey on Hallucination in Large Language Mod- els. arXiv:2309.01219. Model Past l as Context Past l/2 and Next l/2 as Context Context Retrieval- Advanced RAG BLEU COMET BLEU COMET BLEU COMET Aya23 8B 6.38 0.61 6.42 0.61 6.77 0.64 Gemma2 9B 6.56...

    arXiv

  9. [30]

    Vincent, S.; Prescott, C.; Bayliss, C.; Oakley, C.; and Scarton, C

    Curran Associates, Inc. Vincent, S.; Prescott, C.; Bayliss, C.; Oakley, C.; and Scarton, C. 2024a. A Case Study on Contextual Ma- chine Translation in a Professional Scenario of Subtitling. arXiv:2407.00108. Vincent, S.; Sumner, R.; Dowek, A.; Prescott, C.; Pre- ston, E.; Bayl...

    2024 arXiv

  10. [2014]

    In Moschitti, A.; Pang, B.; and Daelemans, W., eds., Proceedings of the 2014 Confer- ence on Empirical Methods in Natural Language Processing (EMNLP), 1724–1734

    Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation. In Moschitti, A.; Pang, B.; and Daelemans, W., eds., Proceedings of the 2014 Confer- ence on Empirical Methods in Natural Language Processing (EMNLP), 1724–1734. Doha, Qatar: Associa...

    2014

  11. [2015]

    In Bengio, Y .; and LeCun, Y ., eds., 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceed- ings

    Neural Ma- chine Translation by Jointly Learning to Align and Trans- late. In Bengio, Y .; and LeCun, Y ., eds., 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceed- ings. Brown, T. B.; Mann, B.; Ryde...

    2015

  12. [2016]

    CoRR, abs/1609.08144

    Google’s Neural Machine Translation Sys- tem: Bridging the Gap between Human and Machine Trans- lation. CoRR, abs/1609.08144. Yao, B.; Jiang, M.; Yang, D.; and Hu, J

    arXiv

  13. [2019]

    arXiv:1909.05362

    Problems with automating translation of movie/TV show subtitles. arXiv:1909.05362. Jiang, A. Q.; Sablayrolles, A.; Mensch, A.; Bamford, C.; Chaplot, D. S.; de las Casas, D.; Bressand, F.; Lengyel, G.; Lample, G.; Saulnier, L.; Lavaud, L. R.; Lachaux, M.-A.; Stock, P.; Scao, T....

    1909 arXiv

  14. [2020]

    arXiv:2005.14165

    Language Models are Few-Shot Learners. arXiv:2005.14165. Cai, W.; Jiang, J.; Wang, F.; Tang, J.; Kim, S.; and Huang, J

    2005 arXiv

  15. [2021]

    arXiv:2005.11401

    Retrieval- Augmented Generation for Knowledge-Intensive NLP Tasks. arXiv:2005.11401. Li, S.; Chen, J.; Yuan, S.; Wu, X.; Yang, H.; Tao, S.; and Xiao, Y

    2005 arXiv

  16. [2022]

    arXiv:2207.06300

    Re2G: Retrieve, Rerank, Gen- erate. arXiv:2207.06300. Gupta, P.; Sharma, M.; Pitale, K.; and Kumar, K

    arXiv

  17. [2023]

    In Rogers, A.; Boyd-Graber, J.; and Okazaki, N., eds., Findings of the Association for Computa- tional Linguistics: ACL 2023, 8857–8873

    In-context Examples Selection for Machine Translation. In Rogers, A.; Boyd-Graber, J.; and Okazaki, N., eds., Findings of the Association for Computa- tional Linguistics: ACL 2023, 8857–8873. Toronto, Canada: Association for Computational Linguistics. Aryabumi, V .; Dang, J.; ...

    2023

  18. [2024]

    arXiv:2405.15032

    Aya 23: Open Weight Releases to Further Multilingual Progress. arXiv:2405.15032. Bahdanau, D.; Cho, K.; and Bengio, Y

    arXiv

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.