Pith. sign in

REVIEW 4 major objections 4 minor 34 references

Using External knowledge to Enhanced PLM for Semantic Matching

T0 review · 4 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

Pith's one-line read The paper claims that calibrating BERT's self-attention with WordNet-derived lexical relation priors yields consistent accuracy gains in semantic matching, with average improvements of 1.66 points over BERT-base and 2.31 points over…

desk verdict Incomplete draft: the central knowledge-prior mapping into self-attention is never specified, so the reported gains cannot be evaluated. read the letter →

arxiv 2505.06605 v1 pith:CDODAD5D submitted 2025-05-10 cs.CL

classification cs.CL
keywords semanticrelevancemodelingexternalknowledgeattentioncalibrationadaptivefusionWordNetrelationspre-trainedlanguagemodelsnaturalinferenceparaphraseidentification
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 argues that pretrained language models cannot reliably learn all the lexical knowledge needed for semantic matching from co-occurrence data alone, and that injecting structured lexical relations from WordNet into BERT's attention improves performance. It proposes a knowledge-aware co-attention that builds a prior matrix from synonym, antonym, hypernym, and hyponym links, injects that prior into self-attention, and uses an adaptive gated fusion to combine semantic and knowledge signals. On ten public datasets, the authors report average accuracy gains of 1.66% over BERT-base and 2.31% over BERT-large on six GLUE tasks, with additional gains on SNLI, SciTail, SICK, and TwitterURL. If the claim holds, this offers a modest but consistent way to make PLMs more sensitive to fine-grained semantic differences without changing pretraining, including better resistance to antonym swaps and similar lexical perturbations.

What carries the argument

The central object is the knowledge prior matrix $K_{\mathrm{prior}}\in\mathbb{R}^{L\times L}$, built from WordNet relation indicators that boost the co-attention scores between words in the two sentences. This matrix is multiplied element-wise into the PLM's scaled dot-product attention scores, producing a knowledge-modulated attention output parallel to the standard semantic output. The second key machinery is the adaptive fusion module: a cross-attention step where each signal gathers context from the other, a fusion gate $g_{\mathrm{fuse}}$ that balances the two contributions, and a filtration gate $g_{\mathrm{filter}}$ that scales the combined signal before it replaces the self-attention output. Together these carry the argument that explicit lexical relations can sharpen attention alignment and then be selectively trusted.

What would settle it

Replace the WordNet-based prior matrix with a randomly permuted matrix of the same shape, keeping the architecture and training unchanged; if the accuracy gains over BERT persist, the reported improvements come from the added capacity of the fusion network, not from lexical knowledge. A direct implementation also requires a concrete rule mapping the $m\times n$ relation matrix to the $L\times L$ self-attention positions, and without that rule Equation (5) cannot be executed.

Watch

Extended reading notes

Core claim

The central claim is that calibrating BERT's self-attention with an external lexical-knowledge prior improves semantic relevance modeling beyond what BERT can learn from text alone. The authors construct a co-attention matrix over the two input sentences, augment each pairwise similarity score with an indicator of whether the word pair is linked in WordNet by synonymy, antonymy, hypernymy, or hyponymy, and average the resulting attention weights to form a prior matrix $K_{\mathrm{prior}}$ that is multiplied element-wise into the scaled dot-product attention scores in each head. A two-stage adaptive fusion then lets the semantic and knowledge-modulated attention outputs adjust each other before a gating network decides how much of the fused signal to keep, yielding a final self-attention output that can fall back on the original semantic representation when the external signal is noisy. The paper reports consistent accuracy improvements over BERT on all evaluated tasks, with the largest relative gains on contradiction-heavy and lexical-variation cases such as swapped antonyms.

Load-bearing premise

The load-bearing premise is that the matrix of WordNet relations computed between words of the two sentences can be rearranged into the square attention matrix inside the pretrained model; the paper assumes this formatting rather than defining it.

Editorial extensions

If this is right

  • On the six GLUE sentence-pair tasks, the WordNet prior raises average accuracy by 1.66 percentage points over BERT-base and 2.31 points over BERT-large, with the largest single-task gain on QQP.
  • Across ten datasets, the method averages higher accuracy than SemBERT, UERBERT, and SyntaxBERT, indicating lexical-relation priors can complement syntax- and semantic-role-based knowledge infusion.
  • In robustness tests, the model gains roughly 10 points over the best baseline on swapped-antonym pairs in QQP, showing the prior sharpens contradiction detection.
  • The case studies show the filtration gate sometimes heavily weights the fused knowledge signal and sometimes nearly ignores it, so the adaptive gate is what makes noisy external knowledge safe to use.

Reading between the lines

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

  • Inference: because the injection is formulated as an additive prior on attention scores, the same mechanism could absorb other discrete lexical resources, such as domain ontologies or medical terminologies, without retraining the PLM; the paper only demonstrates WordNet relations.
  • Inference: the reported gate values in the case studies suggest the model is learnably deciding when to trust external knowledge; one could test this directly by measuring whether gate values correlate with WordNet coverage of the word pairs in each input.
  • Inference: if the $m\times n$ to $L\times L$ reshaping is resolved, the method could be applied to single-sentence tasks by treating the second input as the same sentence, though the paper does not explore this.
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 / 4 minor

Summary. The paper proposes augmenting pre-trained language models (BERT-base and BERT-large) for semantic matching with an external lexical knowledge prior derived from WordNet. The method introduces a knowledge-aware co-attention mechanism, injects a prior matrix into self-attention via elementwise multiplication, and applies an adaptive gated fusion module to combine semantic and knowledge-modulated signals. Experiments are reported on six GLUE tasks and four additional datasets (SNLI, SciTail, SICK, TwitterURL), plus robustness evaluations using TextFlint, with claimed average accuracy gains of 1.66% over BERT-base and 2.31% over BERT-large, together with improvements over prior knowledge-enhanced baselines such as SemBERT, UERBERT, and SyntaxBERT.

Significance. If the proposed mechanism were precisely specified and reproducible, the idea of calibrating attention with explicit lexical relations (synonymy, antonymy, hypernymy) could be a useful contribution to semantic matching and model interpretability. The experimental scope is broad, including 10 datasets and robustness transformations, which is commendable. However, the manuscript's central construction is not concretely defined: the knowledge prior is never mapped to the self-attention space, and the 'external' prior is actually derived from the PLM's own attention weights. The paper also contains explicit 'Self-correction' and 'Please ensure' notes indicating unresolved equations and references. Consequently, the reported gains cannot currently be attributed to the proposed method, and the paper is not in a publishable state.

major comments (4)
  1. [§3.2, Eqs. (3)–(5)] The knowledge prior K_prior is defined only as an m×n co-attention matrix via K_prior[i,j] = (ωA_ij + ωB_ij)/2, where ωA and ωB are computed from the augmented scores s_ij. The text then states 'we assume Kprior is appropriately formatted' and 'Let's assume Kprior ∈ R^{L×L}', but no mapping is given from the (m,n) word-pair indices to the L positions of the concatenated BERT input. There is no treatment of [CLS], [SEP], or padding tokens, and no description of how the m×n matrix is expanded or aligned to the L×L self-attention score matrix. As a result, Eq. (5)'s elementwise multiplication QK^T ⊙ K_prior is undefined in any concrete implementation; this is the load-bearing step of the proposed method.
  2. [§3.2, Eqs. (1)–(5)] The 'external' knowledge prior is not an external fixed matrix. K_prior is the average of the model's own attention weights ωA_ij and ωB_ij, which are computed from s_ij = (h_i^A)^T h_j^B + γ I(k_ij). Thus K_prior depends on the contextualized hidden states h_i^A, h_j^B and on the same knowledge signal G(k_ij) that it is supposed to inject. The injection is therefore circular: the calibration target and the injected prior both derive from the identical augmented scores. Moreover, when no WordNet relation is found, I(k_ij)=0 and K_prior degenerates to the model's ordinary dot-product soft-alignment weights, so the 'knowledge' component does not act as an independent prior at all.
  3. [§3.4, Eqs. (6)–(8)] The adaptive fusion module is not fully specified. In Eq. (6), the ⊕ operator is described as 'concatenation then linear layer', but the dimensions of the operand WOknw O^T_knw ⊕ (W_osem o_sem_i + b_osem) are not defined, and the text inserts 'Ensure weight dimensions match' as an instruction rather than a specification. Further, Eq. (8) defines y_i solely as a gated re-scaling of u_i (y_i = g_filter ⊙ tanh(W_y u_i)), with no direct dependence on o_sem_i except through the gate; yet §5.3 interprets low filter-gate values as the model 'relying more on the original semantic representation (o_sem_i, implicitly weighted higher when g_filter is low)'. This interpretation is inconsistent with the equation as written.
  4. [§5.1, §5.2, §3.4] The manuscript contains explicit non-final notes that undermine its status as a completed submission: §3.4 includes a 'Self-correction' passage saying the original equations were 'overly complex/potentially incorrect' and that the replacements should be assumed, while §5.1 and §5.2 contain 'Please ensure this reference points to the correct table' and 'details assumed to be in Table 2 - ensure this ref is correct'. These passages indicate unresolved equations and references, making it impossible to verify the experimental claims or reproduce the method from the text.
minor comments (4)
  1. [Table 3] Several entries are missing separators between numbers, e.g., '87.5085.48', '50.2083.98', and '60.89/61.3783.23/83.19', making the table difficult to read and some values ambiguous.
  2. [Table 4] The table formatting is broken: fields such as 'label:1 label:0filter gate:0.93label:0' are concatenated without spaces or separators, and the label semantics (what 0 and 1 denote) are not stated in the text.
  3. [Abstract and title] The title contains a grammatical error ('Using External knowledge to Enhanced PLM'), and the abstract repeats informal phrasing such as 'state-ofthe-art'; the paper would benefit from a careful language edit.
  4. [References] Reference [10] is cited in the text as 'UERBERT' but the reference title is 'Using prior knowledge to guide bert's attention in semantic textual matching tasks'; the naming should be made consistent.

Circularity Check

1 steps flagged · score 6.0 of 10

The 'external knowledge prior' is defined as the model's own co-attention weights, making the Eq. (5) calibration a self-modulation.

  1. self definitional [Section 3.2 (Eqs. 3-4) and Section 3.3 (Eq. 5)]
    "We can define an aggregate prior knowledge matrix Kprior∈ Rm×n based on these weights, for example, by averaging: Kprior[i,j ] = (ωA ij +ωB ij )/2. ... For integration into self-attention later, we assume Kprior is appropriately formatted ... Let’s assume Kprior∈ RL×L ... Scoresknw = QKT⊙ Kprior√dk"

    The matrix called the 'external knowledge prior' is defined as the average of the co-attention weights ωA and ωB (Eqs. 3-4), which are themselves softmax functions of the augmented scores s_ij = (h_i^A)^T h_j^B + γI(k_ij). Thus Kprior is not an external fixed prior; it is a function of the PLM's own hidden states and of the same knowledge signal it is supposed to inject. Substituting into Eq. (5) gives Scores_knw = QK^T ⊙ f(QK^T, γI), so the 'calibration' is a self-modulation of the model's own attention by a softmaxed version of itself, not an injection of external knowledge. When no WordNet relation exists, Kprior degenerates to softmax of dot-product similarities, a re-encoding of the model's own representations.

full rationale

The paper's empirical benchmarking is against external public datasets, and there is no load-bearing self-citation: references to papers co-authored by Min Li (e.g., [23]) appear only in robustness-related related work and do not support the central mechanism. However, the central construction of the 'knowledge prior' is self-definitional. Section 3.2 defines Kprior as the average of the co-attention weights (Eqs. 3-4) that are already computed from the augmented scores s_ij containing the WordNet indicator. Section 3.3 then multiplies self-attention scores by this Kprior (Eq. 5). The 'prior' is therefore not an external fixed matrix; it is a function of the same hidden states and knowledge signal it is supposed to inject, so the calibration reduces to a self-modulation of the model's own attention. In addition, the missing m×n to L×L mapping ('we assume Kprior is appropriately formatted') makes Eq. (5) unimplementable as specified, a correctness risk separate from circularity. On balance, the central mechanism is partially circular: the external knowledge enters only inside the softmax that defines the 'prior,' so the claimed attribution of gains to external knowledge is not forced by the construction.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The central claim depends on two hand-tuned injection knobs (γ and the K_prior formatting assumption) and on the standard assumption that WordNet relations transfer to semantic matching without sense disambiguation. No new scientific entities are introduced. The ad hoc formatting assumption for K_prior is the most serious ledger entry because the model cannot be implemented without it.

free parameters (4)
  • gamma (γ) = not reported
    Scaling hyperparameter for the knowledge indicator in Eqs (1)-(2). It controls the strength of the external-knowledge boost to attention scores, and the paper only calls it 'tunable' without giving the chosen value.
  • L2FullRatio = 0.9e-5
    Maximum L2 regularization ratio in Eq (9), set by hand and used in all experiments.
  • L2FullStep = 100,000
    Training step at which the maximum L2 ratio is applied; hand-chosen in Section 4.3.
  • sequence length cutoffs = 48 for MNLI, 32 for SNLI, 24 for Quora
    Hard sequence cutoffs in Section 4.3, introduced ad hoc and likely affecting performance on those datasets.
assumptions (3)
  • domain assumption WordNet lexical relations (synonymy, antonymy, hypernymy, hyponymy) between surface word forms are reliable indicators of semantic relevance and need no word-sense disambiguation.
    Section 3.1 builds the entire knowledge signal on these relations using raw word forms; no word-sense disambiguation or contextual filtering is described.
  • domain assumption A scalar indicator of relation existence is sufficient to calibrate attention.
    Eq (2) uses G(k_ij)=γI(k_ij); the authors call it 'simple yet effective in practice' without supporting analysis.
  • ad hoc to paper The co-attention prior K_prior can be mapped into the self-attention score space of the PLM.
    Section 3.2 states 'we assume Kprior is appropriately formatted' for the L×L self-attention layer. This mapping is never defined.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Using External knowledge to Enhanced PLM for Semantic Matching." pith.science (2026). https://pith.science/paper/CDODAD5D

@misc{pith2026250506605,
  author       = {Pith},
  title        = {Pith review of: Using External knowledge to Enhanced PLM for Semantic Matching},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CDODAD5D}},
  note         = {Machine review of arXiv:2505.06605}
}
read the original abstract

Modeling semantic relevance has always been a challenging and critical task in natural language processing. In recent years, with the emergence of massive amounts of annotated data, it has become feasible to train complex models, such as neural network-based reasoning models. These models have shown excellent performance in practical applications and have achieved the current state-ofthe-art performance. However, even with such large-scale annotated data, we still need to think: Can machines learn all the knowledge necessary to perform semantic relevance detection tasks based on this data alone? If not, how can neural network-based models incorporate external knowledge into themselves, and how can relevance detection models be constructed to make full use of external knowledge? In this paper, we use external knowledge to enhance the pre-trained semantic relevance discrimination model. Experimental results on 10 public datasets show that our method achieves consistent improvements in performance compared to the baseline model.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 20 canonical work pages

  1. [30]

    Transferring from formal newswire domain with hypernet for twitter pos tagging,

    T. Gui, Q. Zhang, J. Gong, M. Peng, D. Liang, K. Ding, and X.-J. Huang, “Transferring from formal newswire domain with hypernet for twitter pos tagging,” in Proceedings of the 2018 conference on empirical methods in natural language processing, 2018, pp. 2540–2549

  2. [1]

    Supervised learning of universal sentence representations from natural language inference data,

    A. Conneau, D. Kiela, H. Schwenk, L. Barrault, and A. Bordes, “Supervised learning of universal sentence representations from natural language inference data,”

  3. [2]

    Wordnet: a lexical database for english,

    G. A. Miller, “Wordnet: a lexical database for english,”Communications of the ACM, vol. 38, no. 11, pp. 39–41, 1995

  4. [3]

    A compare-propagate architecture with alignment factor- ization for natural language inference,

    Y . Tay, L. A. Tuan, and S. C. Hui, “A compare-propagate architecture with alignment factor- ization for natural language inference,”, 2017

  5. [4]

    Syntax-bert: Improving pre-trained transformers with syntax trees,

    J. Bai, Y . Wang, Y . Chen, Y . Yang, J. Bai, J. Yu, and Y . Tong, “Syntax-bert: Improving pre-trained transformers with syntax trees,” arXiv preprint arXiv:2103.04350

  6. [5]

    A fast unified model for parsing and sentence understanding,

    S. R. Bowman, J. Gauthier, A. Rastogi, R. Gupta, C. D. Manning, and C. Potts, “A fast unified model for parsing and sentence understanding,”, 2016

  7. [6]

    Asynchronous deep interaction network for natural language inference,

    D. Liang, F. Zhang, Q. Zhang, and X.-J. Huang, “Asynchronous deep interaction network for natural language inference,” in Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), 2019, pp. 2692–2700

  8. [7]

    Enhanced lstm for natural language inference,

    Q. Chen, X. Zhu, Z. Ling, S. Wei, H. Jiang, and D. Inkpen, “Enhanced lstm for natural language inference,” arXiv preprint arXiv:1609.06038, 2016

Show all 34 references
  1. [8]

    Bert: Pre-training of deep bidirectional transformers for language understanding,

    J. Devlin, M.-W. Chang, K. Lee, and K. Toutanova, “Bert: Pre-training of deep bidirectional transformers for language understanding,” , 2018. Using External knowledge to Enhanced PLM for Semantic Matching 13

  2. [9]

    Semantics-aware bert for language understanding,

    Z. Zhang, Y . Wu, H. Zhao, Z. Li, S. Zhang, X. Zhou, and X. Zhou, “Semantics-aware bert for language understanding,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 34, 2020, pp. 9628–9635

  3. [10]

    Using prior knowledge to guide bert’s attention in semantic textual matching tasks,

    T. Xia, Y . Wang, Y . Tian, and Y . Chang, “Using prior knowledge to guide bert’s attention in semantic textual matching tasks,” in WWW 2021, 2021, pp. 2466–2475

  4. [12]

    Adaptive multi-attention network incorporating answer information for duplicate question detection,

    D. Liang, F. Zhang, W. Zhang, Q. Zhang, J. Fu, M. Peng, T. Gui, and X. Huang, “Adaptive multi-attention network incorporating answer information for duplicate question detection,” in Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Info...

  5. [13]

    DABERT: Dual attention enhanced BERT for semantic matching,

    S. Wang, D. Liang, J. Song, Y . Li, and W. Wu, “DABERT: Dual attention enhanced BERT for semantic matching,” in Proceedings of the 29th International Conference on Computational Linguistics, Oct. 2022

  6. [14]

    Dual path modeling for semantic matching by perceiving subtle conflicts,

    C. Xue, D. Liang, S. Wang, J. Zhang, and W. Wu, “Dual path modeling for semantic matching by perceiving subtle conflicts,” in ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2023

  7. [15]

    Time-aware multiway adaptive fusion network for temporal knowledge graph question answering,

    Y . Liu, D. Liang, F. Fang, S. Wang, W. Wu, and R. Jiang, “Time-aware multiway adaptive fusion network for temporal knowledge graph question answering,” in ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2023, pp. 1–5

  8. [16]

    Enhanced attentive convolutional neural networks for sentence pair modeling,

    S. Xu, E. Shijia, and Y . Xiang, “Enhanced attentive convolutional neural networks for sentence pair modeling,” Expert Systems with Applications, 2020

  9. [17]

    Local and global: temporal question answering via information fusion,

    Y . Liu, M. L. Di Liang, F. Giunchiglia, X. Li, S. Wang, W. Wu, L. Huang, X. Feng, and R. Guan, “Local and global: temporal question answering via information fusion,” inProceedings of the Thirty-Second International Joint Conference on Artificial Intelligence, 2023, pp. 5141–5149

  10. [18]

    Cqg: A simple and effective controlled generation framework for multi-hop question generation,

    Z. Fei, Q. Zhang, T. Gui, D. Liang, S. Wang, W. Wu, and X.-J. Huang, “Cqg: A simple and effective controlled generation framework for multi-hop question generation,” in Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Pape...

  11. [19]

    Distilling structured knowledge for text-based relational reasoning,

    J. Dong, M.-A. Rondeau, and W. L. Hamilton, “Distilling structured knowledge for text-based relational reasoning,” in EMNLP, 2020, pp. 6782–6791

  12. [20]

    Charbert: character-aware pre-trained language model,

    W. Ma, Y . Cui, C. Si, T. Liu, S. Wang, and G. Hu, “Charbert: character-aware pre-trained language model,” arXiv preprint arXiv:2011.01513, 2020

  13. [21]

    Question calibration and multi-hop modeling for temporal question answering,

    C. Xue, D. Liang, P. Wang, and J. Zhang, “Question calibration and multi-hop modeling for temporal question answering,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 38, 2024, pp. 19 332–19 340

  14. [22]

    Learning natural language inference using bidirectional lstm model and inner-attention,

    Y . Liu, C. Sun, L. Lin, and X. Wang, “Learning natural language inference using bidirectional lstm model and inner-attention,”, 2016

  15. [23]

    Resolving word vagueness with scenario-guided adapter for natural language inference,

    Y . Liu, M. Li, D. Liang, X. Li, F. Giunchiglia, L. Huang, X. Feng, and R. Guan, “Resolving word vagueness with scenario-guided adapter for natural language inference,” arXiv preprint arXiv:2405.12434, 2024

  16. [24]

    Local and global: Text matching via syntax graph calibration,

    L. Li, Q. Liao, M. Lai, D. Liang, and S. Liang, “Local and global: Text matching via syntax graph calibration,” inICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2024, pp. 11 571–11 575

  17. [25]

    Glue: A multi-task benchmark and analysis platform for natural language understanding,

    A. Wang, A. Singh, J. Michael, F. Hill, O. Levy, and S. R. Bowman, “Glue: A multi-task benchmark and analysis platform for natural language understanding,”

  18. [26]

    A large annotated corpus for learning natural language inference,

    S. R. Bowman, G. Angeli, C. Potts, and C. D. Manning, “A large annotated corpus for learning natural language inference,” arXiv preprint arXiv:1508.05326, 2015. 14 Li et al

  19. [27]

    Robust lottery tickets for pre-trained language models,

    R. Zheng, R. Bao, Y . Zhou, D. Liang, S. Wang, W. Wu, T. Gui, Q. Zhang, and X. Huang, “Robust lottery tickets for pre-trained language models,”arXiv preprint arXiv:2211.03013, 2022

  20. [28]

    Searching for optimal subword tokenization in cross-domain ner,

    R. Ma, Y . Tan, X. Zhou, X. Chen, D. Liang, S. Wang, W. Wu, T. Gui, and Q. Zhang, “Searching for optimal subword tokenization in cross-domain ner,” arXiv preprint arXiv:2206.03352, 2022

  21. [29]

    Improving semantic matching through dependency-enhanced pre-trained model with adaptive fusion,

    J. Song, D. Liang, R. Li, Y . Li, S. Wang, M. Peng, W. Wu, and Y . Yu, “Improving semantic matching through dependency-enhanced pre-trained model with adaptive fusion,” in Findings of the Association for Computational Linguistics: EMNLP 2022. Abu Dhabi, United Arab Emirates: A...

  22. [31]

    Unleashing potential of evidence in knowledge-intensive dialogue generation,

    X. Wu, J. Yang, T. Li, S. Zhang, Y . Du, L. Chai, D. Liang, and Z. Li, “Unleashing potential of evidence in knowledge-intensive dialogue generation,” inICASSP 2025-2025 IEEE Interna- tional Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2025, pp. 1–5

  23. [32]

    Attention is all you need,

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention is all you need,” in Advances in neural information processing systems, 2017, pp. 5998–6008

  24. [33]

    Textflint: Unified multilingual robustness evaluation toolkit for natural language processing,

    T. Gui, X. Wang, Q. Zhang, Q. Liu, Y . Zou, X. Zhou, R. Zheng, C. Zhang, Q. Wu, J. Yeet al., “Textflint: Unified multilingual robustness evaluation toolkit for natural language processing,” arXiv preprint arXiv:2103.11441, 2021

  25. [34]

    Tablebench: A comprehensive and complex benchmark for table question answering,

    X. Wu, J. Yang, L. Chai, G. Zhang, J. Liu, X. Du, D. Liang, D. Shu, X. Cheng, T. Sunet al., “Tablebench: A comprehensive and complex benchmark for table question answering,”arXiv preprint arXiv:2408.09174, 2024

  26. [35]

    Comateformer: Combined attention transformer for semantic sentence matching,

    B. Li, D. Liang, and Z. Zhang, “Comateformer: Combined attention transformer for semantic sentence matching,” arXiv preprint arXiv:2412.07220, 2024

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.