Pith. sign in

REVIEW 4 major objections 5 minor 62 references

Can Hallucinations Help? Boosting LLMs for Drug Discovery

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · 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 LLM hallucinations can be harnessed: adding hallucinated molecule descriptions to prompts improves molecular property prediction, sometimes beating factual chemical databases like PubChem.

desk verdict Interesting and transparent, but the headline gains are single-run point estimates on tiny test sets and the abstract overstates what the paper's own tables show. read the letter →

arxiv 2501.13824 v2 pith:YQ5N4BXO submitted 2025-01-23 cs.CL cs.AI

Shuzhou Yuan , Zhan Qu , Ashish Yashwanth Kangen , Michael Färber This is my paper · ORCID
classification cs.CLcs.AI
keywords hallucinationmoleculepropertypredictiondrugdiscoverySMILESlargelanguagemodelsimplicitcounterfactualsROC-AUCMolT5
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 overturn the usual assumption that hallucinations are always bad: it asks whether the plausible-but-wrong molecule descriptions that LLMs generate can actually help the same models predict molecular properties, such as whether a compound inhibits HIV or crosses the blood-brain barrier. The authors prompt seven instruction-tuned LLMs to describe SMILES strings, then feed those descriptions back into the models as extra context for binary property prediction across five MoleculeNet datasets. They report that hallucinated context often beats two clean baselines—SMILES-only prompts and MolT5 reference descriptions—and in some cases even beats structured PubChem metadata. The practical stake is that hallucinated text, treated as a kind of implicit counterfactual, could be a cheap source of inductive bias for early-stage drug discovery where structured knowledge is sparse.

What carries the argument

The load-bearing mechanism is the prompt template [SMILES][Description][Instruct], where a binary yes/no prediction is decoded by the highest-probability token over {Yes, No}. Hallucination is measured by the HHEM-2.1-Open model, which scores alignment between each LLM-generated description and a MolT5 reference description; low scores indicate high hallucination. A four-category typology—structural misdescription, functional hallucination, analogical hallucination, and generic fluff—is used to label the 18,000+ beneficial hallucinations, and structural misdescription is identified as the most impactful type.

What would settle it

Re-run the five benchmark tasks after replacing each hallucinated description with either randomly sampled chemistry-flavored text or with verified descriptions from a trusted database; if the gains persist with random text, the effect is not about hallucination content but merely about adding extra tokens, and the implicit-counterfactual explanation would be unsupported. Alternatively, verify the reported structural misdescriptions against computed molecular structure, and check whether they are actually incorrect in most cases.

Watch

Extended reading notes

Core claim

The central discovery is that hallucinated descriptions, interpreted as implicit counterfactuals, can serve as a useful inductive bias for LLM molecular property prediction. Concretely, Falcon3-Mamba-7B achieves an average ROC-AUC 8.22% higher than the PubChem baseline when given GPT-4o-generated descriptions, and Llama-3.1-8B improves over SMILES by 15.80% and over MolT5 by 11.23% when given Llama-3 descriptions. The authors further collect 18,872 cases where hallucinations improve predictions over all baselines for Falcon3-Mamba-7B and find that 85% of these beneficial hallucinations are structural misdescriptions—incorrect but chemically plausible statements about atoms, bonds, or substructures—rather than functional, analogical, or vague text. Larger models extract more benefit from hallucinated context, while sampling temperature affects factuality but has only a small, non-monotonic effect on downstream performance.

Load-bearing premise

The paper equates hallucination with divergence from MolT5 reference descriptions as scored by HHEM-2.1-Open, but MolT5 is itself a generated model rather than verified ground-truth chemistry, so what is called hallucination is really dissimilarity to one particular reference text.

Editorial extensions

If this is right

  • Adding LLM-generated description text to SMILES prompts can raise ROC-AUC by double digits for some models, beating the SMILES-only and MolT5 baselines.
  • The source of the hallucination matters: GPT-4o-generated descriptions give the most consistent gains across all seven predictor models.
  • Among hallucination types, structural misdescriptions drive most of the benefit, suggesting that factually wrong but chemically plausible structural claims help LLM decision-making.
  • Larger models (up to 8B parameters) extract more signal from hallucinated text, while generation temperature between 0.1 and 0.9 has limited influence on downstream accuracy.
  • Hallucinated context can, in some configurations, outperform a gold-standard rule-based baseline built from PubChem metadata.

Reading between the lines

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

  • The reported gains might not require genuine hallucination: a controlled experiment substituting random chemistry-flavored text for the hallucinated descriptions would reveal whether it is the content or merely the presence of extra text that helps.
  • Because the paper measures hallucination as divergence from MolT5 rather than from verified structure, the practical recipe may transfer to any domain where a reference description model exists, not just chemistry.
  • The dominance of structural misdescriptions suggests a testable extension: deliberately perturbing SMILES strings (for example, swapping atoms or bonds) to generate targeted counterfactual descriptions could reproduce the benefit in a more controlled and interpretable way.
  • If the effect is robust, hallucination could be treated as a tunable source of synthetic data for scientific prediction tasks, with temperature and model size as knobs rather than things to suppress.
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

4 major / 5 minor

Summary. The paper asks whether hallucinated natural-language descriptions of molecules can improve LLM-based molecular property prediction. For each SMILES string, the authors prompt one of seven LLMs to generate a free-text description, append it to a binary classification prompt, and compare the resulting ROC-AUC against three non-LLM description settings: SMILES-only, MolT5 reference descriptions, and rule-based PubChem metadata. Across five MoleculeNet datasets, they report that some models benefit substantially, e.g., Falcon3-Mamba-7B improving by 8.22% over PubChem with GPT-4o descriptions and Llama-3.1-8B improving by 15.80% over SMILES with Llama-3 descriptions. The paper also proposes a four-type hallucination taxonomy, uses an external LLM to label beneficial hallucinations, and reports ablations over model size and generation temperature.

Significance. The question is timely and the experimental setup is broad: seven instruction-tuned models, five datasets, three baselines, and a separate evaluation of which hallucination source helps most. The core comparison against SMILES, MolT5, and PubChem involves no fitted parameters and is therefore not circular, and the hallucination taxonomy with a human-agreement check is a useful contribution. If the performance differences survive uncertainty quantification and the hallucination measurement is validated against actual molecular structure, the result would be a valuable empirical finding for prompt engineering in scientific LLM use. As it stands, the evidence is suggestive rather than conclusive, because the headline numbers are single-run point estimates, the main table is selected after seeing performance, and the term 'hallucination' is attached to divergence from a neural reference rather than to verified chemical inaccuracy.

major comments (4)
  1. [Table 5 / Table 7] All reported ROC-AUC values are single-run point estimates with no confidence intervals, bootstrap estimates, DeLong tests, or repeated seeds. This matters because Table 5 reports very small test sets: BBBP has 205 instances, Clintox has 148 instances with only 11 positives, and SIDER has 143 instances. For example, Llama-3.1-8B's Clintox score moves from 35.30 with SMILES to 68.08 with Llama-3 descriptions; with 11 positives, a few re-ranked examples can produce this difference, and the standard error of AUC is large. Since the headline gains (e.g., +15.80% over SMILES for Llama-3.1-8B and +8.22% over PubChem for Falcon3-Mamba-7B) are averages over five datasets, one noisy small dataset can drive the average. The paper should provide per-dataset uncertainty estimates, significance tests, or repeated runs before claiming that hallucinations 'significantly improve' predictive accuracy. In addition, SIDER and Tox21 use a single label selected for balance, and no sensitivity analysis for that choice is reported.
  2. [Table 2 / 'Do Hallucinations Improve LLMs?'] The main results table is curated after seeing performance. Table 2 is captioned 'curated set of descriptions,' and the text states that the hallucination source shown for each model category is the one whose generated hallucinations 'resulted in the greatest average performance gains in ROC-AUC.' Thus the headline comparisons are selected from the full results in Table 7 rather than following a fixed, pre-specified protocol, which can inflate the apparent benefit of hallucinated text. The abstract's claim that 'Falcon3-Mamba-7B outperforms all baselines when hallucinated text is included' is also contradicted by Table 7: with Ministral-generated descriptions, Falcon3-Mamba-7B averages 44.50, below the PubChem baseline of 45.47. The paper should present the full results as primary or clearly label the selected results as best-case, and it should temper the abstract accordingly.
  3. [Hallucination Generation / Figure 1] The definition of hallucination used in the paper is not the quantity that is measured. The Hallucination Generation section defines hallucination as 'any information in the generated text that is not supported by the molecule's actual structure or known properties,' but Figure 1 and Table 6 measure factual consistency with HHEM-2.1-Open relative to MolT5 reference descriptions. MolT5 is a generative model, not ground-truth chemistry, so divergence from MolT5 is not evidence of hallucination against the molecule; a description can diverge from MolT5 and still be chemically accurate. Therefore statements such as 'the majority of generated descriptions contain some degree of hallucination' and the low factual-consistency scores in Figure 1 are not established by the experiment. The authors should either verify descriptions with a structure-aware chemical tool or explicitly reframe the investigation as the effect of descriptions that diverge from a neural reference.
  4. [What Types of Hallucinations Are Most Useful? / Figure 4] The beneficial-hallucination analysis is selection-on-outcome and does not support the causal claim about hallucination types. In 'What Types of Hallucinations Are Most Useful?', beneficial hallucinations are defined as instances that lead to correct predictions with higher confidence than any baseline; Figure 4 then reports that 85% of these beneficial cases are structural misdescriptions. Because the set is selected on the outcome, the type distribution among beneficial cases cannot show that structural misdescriptions are 'the most impactful type' unless it is compared with the type distribution among all generated descriptions or among non-beneficial ones. The annotation quality check is also limited: Fleiss' kappa of 0.57 on 50 instances is moderate, and no reliability is reported for the large-scale Deepseek-R1 annotations. This section should be reworded as a descriptive analysis of selected cases, with appropriate comparisons and uncertainty.
minor comments (5)
  1. [Hallucination Generation / Table 6] The scoring model name is inconsistent: 'HHM-2.1-Open' appears in the Hallucination Generation section and Table 6, while Figure 1 uses 'HHEM-2.1-Open.' The correct name should be used consistently.
  2. [Appendix: Prompt Template] The prompt templates render '[SM ILES]' and '[HALLU CIN AT ION]' with inserted spaces; these appear to be LaTeX line-breaking artifacts and should be corrected.
  3. [Table 4] The HuggingFace link for Llama-3-8B points to the Llama-3.1-8B-Instruct repository; the link should be verified and corrected.
  4. [Effect of Generation Temperature] The text says that hallucination scores increase as temperature increases, but Table 12 is non-monotonic and the 0.9 setting has the lowest HHEM score (6.82). Please reconcile Figure 6 with Table 12 or define explicitly whether 'hallucination score' increases with factual consistency or with divergence.
  5. [Reproducibility] No code, generated descriptions, or data artifacts are made available. Given that the entire analysis depends on prompt templates and generated text, releasing these would substantially improve reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the hallucination-helping claim rests on an independent benchmark comparison.

full rationale

The central claim that LLM-generated molecule descriptions improve ROC-AUC relative to SMILES, MolT5, and PubChem baselines is a direct, parameter-free comparison on fixed MoleculeNet splits. No parameter is fitted to the test outcomes, no predicted quantity is defined in terms of the input descriptions, and the headline gains are not derived from the paper's definition of hallucination. The HHEM-2.1-Open/MolT5 consistency scores are an interpretive proxy for hallucination rather than a fitted ingredient in the AUC comparisons. The beneficial-hallucination analysis selects cases by improved confidence and then reports the type distribution among those cases; this is a post-hoc, selection-on-outcome characterization and a possible base-rate confound, but the type distribution itself is an empirical annotation result, not an equation that reduces to the selection criterion. The absence of significance testing is a statistical robustness concern, not a circularity. Self-citations (Yuan and Faerber 2023; Qu, Gomm, and Färber 2025) appear only in background and related-work discussions and are not load-bearing. No uniqueness theorem or ansatz is imported from the authors' prior work. Thus no circular step meets the quoted-reduction standard; score 0.

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

The paper introduces no new physical or mathematical entities. Its claims rest on benchmark assumptions, a proxy measure of hallucination (divergence from MolT5 via HHEM-2.1), post hoc label selection for two datasets, and an outcome-defined notion of beneficial hallucinations. There are no fitted parameters in the main comparisons, but the description temperature, token limit, label choices, and beneficial-hallucination threshold are hand-chosen settings that the results depend on.

free parameters (4)
  • description generation temperature = 0.6
    Chosen by hand for all LLM description generation; it controls how hallucinated the treatment text is, so the central comparison depends on this setting.
  • max new tokens for descriptions = 256
    Token limit for generated descriptions; restricts description length and may affect the distribution of hallucination types and downstream accuracy.
  • SIDER and Tox21 label selection = reproductive system and breast disorders for SIDER; SR-MMP for Tox21
    One label per multi-label dataset was selected post hoc for the most balanced test distribution; all reported SIDER and Tox21 scores depend on this choice.
  • beneficial hallucination definition = correct prediction with confidence greater than all baselines
    Ad hoc rule used to collect the 18,872-case subset; the subsequent type distribution is conditional on this threshold and does not compare with non-beneficial hallucinations.
assumptions (4)
  • domain assumption HHEM-2.1-Open scores relative to MolT5 reference descriptions measure the degree of hallucination in molecule descriptions.
    Used to support Figure 1 and the claim that most descriptions are hallucinated; MolT5 is itself a generated reference, so divergence is not necessarily factual error.
  • ad hoc to paper The selected SIDER and Tox21 labels are representative of those datasets for evaluating hallucination benefits.
    The labels were chosen for most balanced test distribution, a post hoc choice that shapes average ROC-AUC and limits generalization to the full multitask benchmarks.
  • domain assumption MoleculeNet labels and splits are treated as ground truth.
    Standard benchmark assumption; assay labels contain noise and scaffold/random splits affect difficulty.
  • ad hoc to paper Beneficial hallucinations can be identified as correct predictions with higher confidence than any baseline.
    This selection rule defines the 18,872-case analysis; it guarantees benefit by construction and does not establish which hallucination types cause improvement.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Can Hallucinations Help? Boosting LLMs for Drug Discovery." pith.science (2026). https://pith.science/paper/YQ5N4BXO

@misc{pith2026250113824,
  author       = {Pith},
  title        = {Pith review of: Can Hallucinations Help? Boosting LLMs for Drug Discovery},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YQ5N4BXO}},
  note         = {Machine review of arXiv:2501.13824}
}
read the original abstract

Hallucinations in large language models (LLMs), plausible but factually inaccurate text, are often viewed as undesirable. However, recent work suggests that such outputs may hold creative potential. In this paper, we investigate whether hallucinations can improve LLMs on molecule property prediction, a key task in early-stage drug discovery. We prompt LLMs to generate natural language descriptions from molecular SMILES strings and incorporate these often hallucinated descriptions into downstream classification tasks. Evaluating seven instruction-tuned LLMs across five datasets, we find that hallucinations significantly improve predictive accuracy for some models. Notably, Falcon3-Mamba-7B outperforms all baselines when hallucinated text is included, while hallucinations generated by GPT-4o consistently yield the greatest gains between models. We further identify and categorize over 18,000 beneficial hallucinations, with structural misdescriptions emerging as the most impactful type, suggesting that hallucinated statements about molecular structure may increase model confidence. Ablation studies show that larger models benefit more from hallucinations, while temperature has a limited effect. Our findings challenge conventional views of hallucination as purely problematic and suggest new directions for leveraging hallucinations as a useful signal in scientific modeling tasks like drug discovery.

Figures

Figures reproduced from arXiv: 2501.13824 by the authors.

Figure 1
Figure 1. Factual consistency scores evaluated using the [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Illustration of our method using a sample from the HIV dataset. We first generate hallucinated molecule descriptions [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 4
Figure 4. Distribution of hallucination types contributing to [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figures from the paper (2 more)
Figure 6
Figure 6. Figure 6: Average model performance and hallucination [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 5
Figure 5. Figure 5: Average ROC-AUC improvement for Llama-3 models of different sizes, using hallucinated descriptions. Effect of Generation Temperature Sampling temperature controls the randomness of genera￾tion (Van Koevering and Kleinberg 2024), with higher tem￾peratures producing more…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

62 extracted references · 42 canonical work pages

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  2. [2]

    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 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    V.; Xu, Y.; and Fung, P

    Bang, Y.; Cahyawijaya, S.; Lee, N.; Dai, W.; Su, D.; Wilie, B.; Lovenia, H.; Ji, Z.; Yu, T.; Chung, W.; Do, Q. V.; Xu, Y.; and Fung, P. 2023. A Multitask, Multilingual, Multimodal Evaluation of C hat GPT on Reasoning, Hallucination, and Interactivity. In Park, J. C.; Arase, Y.; Hu, B.; Lu, W.; Wijaya, D.; Purwarianti, A.; and Krisnadhi, A. A., eds., Proce...

    work page 2023

  4. [4]

    Bao, F.; Li, M.; Luo, R.; and Mendelevitch, O. 2024. HHEM-2.1-Open

    work page 2024

  5. [5]

    W.; and Murcko, M

    Bemis, G. W.; and Murcko, M. A. 1996. The properties of known drugs. 1. Molecular frameworks. Journal of medicinal chemistry, 39(15): 2887--2893

    work page 1996

  6. [6]

    A.; MacKnight, R.; Kline, B.; and Gomes, G

    Boiko, D. A.; MacKnight, R.; Kline, B.; and Gomes, G. 2023. Autonomous chemical research with large language models. Nature, 624(7992): 570--578

    work page 2023

  7. [7]

    Byrne, R. M. 2019. Counterfactuals in explainable artificial intelligence (XAI): Evidence from human reasoning. In IJCAI, 6276--6282. California, CA

    work page 2019

  8. [8]

    Chakraborty, C.; Bhattacharya, M.; and Lee, S.-S. 2023. Artificial intelligence enabled ChatGPT and large language models in drug target discovery, drug discovery, and development. Molecular Therapy-Nucleic Acids, 33: 866--868

    work page 2023

Show all 62 references
  1. [9]

    C.; Straitouri, E.; Tsirtsis, S.; and Gomez-Rodriguez, M

    Chatzi, I.; Benz, N. C.; Straitouri, E.; Tsirtsis, S.; and Gomez-Rodriguez, M. 2024. Counterfactual token generation in large language models. arXiv preprint arXiv:2409.17027

    2024 arXiv

  2. [10]

    Chen, H.; and Ding, N. 2023. Probing the `` Creativity '' of Large Language Models: Can models produce divergent semantic association? In Bouamor, H.; Pino, J.; and Bali, K., eds., Findings of the Association for Computational Linguistics: EMNLP 2023, 12881--12888. Singapore: ...

    2023

  3. [11]

    Dhuliawala, S.; Komeili, M.; Xu, J.; Raileanu, R.; Li, X.; Celikyilmaz, A.; and Weston, J. 2024. Chain-of-Verification Reduces Hallucination in Large Language Models. In Ku, L.-W.; Martins, A.; and Srikumar, V., eds., Findings of the Association for Computational Linguistics: ...

    2024

  4. [12]

    Dixit, T.; Paranjape, B.; Hajishirzi, H.; and Zettlemoyer, L. 2022. CORE: A Retrieve-then-Edit Framework for Counterfactual Data Generation. In Findings of the Association for Computational Linguistics: EMNLP 2022, Abu Dhabi, United Arab Emirates, December 7-11, 2022 , 2964--2...

    2022

  5. [13]

    Dubey, A.; Jauhri, A.; Pandey, A.; Kadian, A.; Al-Dahle, A.; Letman, A.; Mathur, A.; Schelten, A.; Yang, A.; Fan, A.; et al. 2024. The llama 3 herd of models. arXiv preprint arXiv:2407.21783

    2024 arXiv

  6. [14]

    Edwards, C.; Lai, T.; Ros, K.; Honke, G.; Cho, K.; and Ji, H. 2022. Translation between Molecules and Natural Language. In Goldberg, Y.; Kozareva, Z.; and Zhang, Y., eds., Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, 375--413. Abu Dha...

    2022

  7. [15]

    Edwards, C.; Wang, Q.; and Ji, H. 2024. Language + Molecules. In Mesgar, M.; and Lo \'a iciga, S., eds., Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics: Tutorial Abstracts, 14--20. St. Julian ' s, Malta: Association ...

    2024

  8. [16]

    Fossi, G.; Boulaimen, Y.; Outemzabet, L.; Jeanray, N.; Gerart, S.; Vachenc, S.; Giemza, J.; and Raieli, S. 2024. SwiftDossier: Tailored Automatic Dossier for Drug Discovery with LLMs and Agents. arXiv preprint arXiv:2409.15817

    2024 arXiv

  9. [17]

    Franceschelli, G.; and Musolesi, M. 2024. On the creativity of large language models. AI & SOCIETY, 1--11

    2024

  10. [18]

    Ganeeva, V.; Sakhovskiy, A.; Khrabrov, K.; Savchenko, A.; Kadurin, A.; and Tutubalina, E. 2024. Lost in Translation: Chemical Language Models and the Misunderstanding of Molecule Structures. In Al-Onaizan, Y.; Bansal, M.; and Chen, Y.-N., eds., Findings of the Association for ...

    2024

  11. [19]

    Ge, Y.; Hua, W.; Mei, K.; ji, j.; Tan, J.; Xu, S.; Li, Z.; and Zhang, Y. 2023. OpenAGI: When LLM Meets Domain Experts. In Advances in Neural Information Processing Systems, volume 36, 5539--5568. Curran Associates, Inc

    2023

  12. [20]

    G \'o mez-Rodr \' guez, C.; and Williams, P. 2023. A Confederacy of Models: a Comprehensive Evaluation of LLM s on Creative Writing. In Bouamor, H.; Pino, J.; and Bali, K., eds., Findings of the Association for Computational Linguistics: EMNLP 2023, 14504--14528. Singapore: As...

    2023

  13. [21]

    Guan, S.; and Wang, G. 2024. Drug discovery and development in the era of artificial intelligence: From machine learning to large language models. Artificial Intelligence Chemistry, 2(1): 100070

    2024

  14. [22]

    Guidotti, R. 2024. Counterfactual explanations and how to find them: literature review and benchmarking. Data Mining and Knowledge Discovery, 38(5): 2770--2824

    2024

  15. [23]

    Guo, D.; Yang, D.; Zhang, H.; Song, J.; Zhang, R.; Xu, R.; Zhu, Q.; Ma, S.; Wang, P.; Bi, X.; et al. 2025. Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning. arXiv preprint arXiv:2501.12948

    2025 arXiv

  16. [24]

    Huang, L.; Yu, W.; Ma, W.; Zhong, W.; Feng, Z.; Wang, H.; Chen, Q.; Peng, W.; Feng, X.; Qin, B.; et al. 2023. A survey on hallucination in large language models: Principles, taxonomy, challenges, and open questions. ACM Transactions on Information Systems

    2023

  17. [25]

    P.; Rees, S.; Kalindjian, S

    Hughes, J. P.; Rees, S.; Kalindjian, S. B.; and Philpott, K. L. 2011. Principles of early drug discovery. British journal of pharmacology, 162(6): 1239--1249

    2011

  18. [26]

    Ji, Z.; Yu, T.; Xu, Y.; Lee, N.; Ishii, E.; and Fung, P. 2023. Towards Mitigating LLM Hallucination via Self Reflection. In Bouamor, H.; Pino, J.; and Bali, K., eds., Findings of the Association for Computational Linguistics: EMNLP 2023, 1827--1843. Singapore: Association for ...

    2023

  19. [27]

    Jiang, X.; Tian, Y.; Hua, F.; Xu, C.; Wang, Y.; and Guo, J. 2024. A Survey on Large Language Model Hallucination via a Creativity Perspective. arXiv:2402.06647

    2024 arXiv

  20. [28]

    Kim, S.; Chen, J.; Cheng, T.; and et al. 2025. PubChem 2025 update. Nucleic Acids Research

    2025

  21. [29]

    Lee, H.-K. 2022. Rethinking creativity: creative industries, AI and everyday creativity. Media, Culture & Society, 44(3): 601--612

    2022

  22. [30]

    Lee, M. 2023. A mathematical investigation of hallucination and creativity in GPT models. Mathematics, 11(10): 2320

    2023

  23. [31]

    Liu, S.; Lu, Y.; Chen, S.; Hu, X.; Zhao, J.; Fu, T.; and Zhao, Y. 2024. DrugAgent: Automating AI-aided Drug Discovery Programming through LLM Multi-Agent Collaboration. arXiv preprint arXiv:2411.15692

    2024 arXiv

  24. [32]

    Liu, S.; Nie, W.; Wang, C.; Lu, J.; Qiao, Z.; Liu, L.; Tang, J.; Xiao, C.; and Anandkumar, A. 2023. Multi-modal molecule structure--text model for text-based retrieval and editing. Nature Machine Intelligence, 5(12): 1447--1457

    2023

  25. [33]

    R.; Subramanian, S.; Mohr, B

    Madani, A.; Krause, B.; Greene, E. R.; Subramanian, S.; Mohr, B. P.; Holton, J. M.; Olmos, J. L.; Xiong, C.; Sun, Z. Z.; Socher, R.; et al. 2023. Large language models generate functional protein sequences across diverse families. Nature Biotechnology, 41(8): 1099--1106

    2023

  26. [34]

    Manakul, P.; Liusie, A.; and Gales, M. 2023. S elf C heck GPT : Zero-Resource Black-Box Hallucination Detection for Generative Large Language Models. In Bouamor, H.; Pino, J.; and Bali, K., eds., Proceedings of the 2023 Conference on Empirical Methods in Natural Language Proce...

    2023

  27. [35]

    Maynez, J.; Narayan, S.; Bohnet, B.; and McDonald, R. 2020. On Faithfulness and Factuality in Abstractive Summarization. In Jurafsky, D.; Chai, J.; Schluter, N.; and Tetreault, J., eds., Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, 1...

    2020

  28. [36]

    Murakumo, K.; Yoshikawa, N.; Rikimaru, K.; Nakamura, S.; Furui, K.; Suzuki, T.; Yamasaki, H.; Nishigaya, Y.; Takagi, Y.; and Ohue, M. 2023. LLM Drug Discovery Challenge: A Contest as a Feasibility Study on the Utilization of Large Language Models in Medicinal Chemistry. In AI ...

    2023

  29. [37]

    A.; and Chakraborty, C

    Pal, S.; Bhattacharya, M.; Islam, M. A.; and Chakraborty, C. 2023. ChatGPT or LLM in next-generation drug discovery and development: pharmaceutical and biotechnology companies can make use of the artificial intelligence-based device for a faster way of drug discovery and devel...

    2023

  30. [38]

    Peeperkorn, M.; Kouwenhoven, T.; Brown, D.; and Jordanous, A. 2024. Is temperature the creativity parameter of large language models? arXiv preprint arXiv:2405.00492

    2024 arXiv

  31. [39]

    A.; Prenkaj, B.; Stilo, G.; and Giannotti, F

    Prado-Romero, M. A.; Prenkaj, B.; Stilo, G.; and Giannotti, F. 2024. A survey on graph counterfactual explanations: definitions, methods, evaluation, and research challenges. ACM Computing Surveys, 56(7): 1--37

    2024

  32. [40]

    Qiu, X.; Wang, Y.; Guo, X.; Zeng, Z.; Yue, Y.; Feng, Y.; and Miao, C. 2024. PairCFR: Enhancing Model Training on Paired Counterfactually Augmented Data through Contrastive Learning. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Vol...

    2024

  33. [41]

    Qu, Z.; Gomm, D.; and F \"a rber, M. 2025. CoDy: Counterfactual Explainers for Dynamic Graphs. In Forty-second International Conference on Machine Learning

    2025

  34. [42]

    Rawte, V.; Chakraborty, S.; Pathak, A.; Sarkar, A.; Tonmoy, S. T. I.; Chadha, A.; Sheth, A.; and Das, A. 2023. The Troubling Emergence of Hallucination in Large Language Models - An Extensive Definition, Quantification, and Prescriptive Remediations. In Bouamor, H.; Pino, J.; ...

    2023

  35. [43]

    Renze, M. 2024. The Effect of Sampling Temperature on Problem Solving in Large Language Models. In Al-Onaizan, Y.; Bansal, M.; and Chen, Y.-N., eds., Findings of the Association for Computational Linguistics: EMNLP 2024, 7346--7356. Miami, Florida, USA: Association for Computa...

    2024

  36. [44]

    Savage, N. 2023. Drug discovery companies are customizing ChatGPT: here’s how. Nat Biotechnol, 41(5): 585--586

    2023

  37. [45]

    Teufel, J.; Leinweber, A.; and Friederich, P. 2025. Improving Counterfactual Truthfulness for Molecular Property Prediction through Uncertainty Quantification. arXiv preprint arXiv:2504.02606

    2025 arXiv

  38. [46]

    Van Koevering, K.; and Kleinberg, J. 2024. How Random is Random? Evaluating the Randomness and Humaness of LLMs' Coin Flips. arXiv preprint arXiv:2406.00092

    2024 arXiv

  39. [47]

    Wang, F. 2024. LightHouse: A Survey of AGI Hallucination. arXiv preprint arXiv:2401.06792

    2024 arXiv

  40. [48]

    Wang, Y.; Qiu, X.; Yue, Y.; Guo, X.; Zeng, Z.; Feng, Y.; and Shen, Z. 2024. A Survey on Natural Language Counterfactual Generation. In Findings of the Association for Computational Linguistics: EMNLP 2024, 4798--4818

    2024

  41. [49]

    Weininger, D. 1988. SMILES, a chemical language and information system. 1. Introduction to methodology and encoding rules. Journal of chemical information and computer sciences, 28(1): 31--36

    1988

  42. [50]

    P.; Seshadri, A.; and White, A

    Wellawatte, G. P.; Seshadri, A.; and White, A. D. 2022. Model agnostic generation of counterfactual explanations for molecules. Chemical science, 13(13): 3697--3705

    2022

  43. [51]

    T.; Heer, J.; and Weld, D

    Wu, T.; Ribeiro, M. T.; Heer, J.; and Weld, D. 2021. Polyjuice: Generating Counterfactuals for Explaining, Evaluating, and Improving Models. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference o...

    2021

  44. [52]

    N.; Gomes, J.; Geniesse, C.; Pappu, A

    Wu, Z.; Ramsundar, B.; Feinberg, E. N.; Gomes, J.; Geniesse, C.; Pappu, A. S.; Leswing, K.; and Pande, V. 2018. MoleculeNet: a benchmark for molecular machine learning. Chemical science, 9(2): 513--530

    2018

  45. [53]

    Yao, Y.; Duan, J.; Xu, K.; Cai, Y.; Sun, Z.; and Zhang, Y. 2024. A survey on large language model (llm) security and privacy: The good, the bad, and the ugly. High-Confidence Computing, 100211

    2024

  46. [54]

    Ye, H.; Liu, T.; Zhang, A.; Hua, W.; and Jia, W. 2023. Cognitive mirage: A review of hallucinations in large language models. arXiv preprint arXiv:2309.06794

    2023 arXiv

  47. [55]

    Yuan, S.; and Faerber, M. 2023. Evaluating Generative Models for Graph-to-Text Generation. In Mitkov, R.; and Angelova, G., eds., Proceedings of the 14th International Conference on Recent Advances in Natural Language Processing, 1256--1264. Varna, Bulgaria: INCOMA Ltd., Shoum...

    2023

  48. [56]

    Zhang, D.; Liu, W.; Tan, Q.; Chen, J.; Yan, H.; Yan, Y.; Li, J.; Huang, W.; Yue, X.; Zhou, D.; et al. 2024 a . Chemllm: A chemical large language model. arXiv preprint arXiv:2402.06852

    2024 arXiv

  49. [57]

    Zhang, H.; Song, Y.; Hou, Z.; Miret, S.; and Liu, B. 2024 b . H oney C omb: A Flexible LLM -Based Agent System for Materials Science. In Al-Onaizan, Y.; Bansal, M.; and Chen, Y.-N., eds., Findings of the Association for Computational Linguistics: EMNLP 2024, 3369--3382. Miami,...

    2024

  50. [58]

    Zhang, K.; Feng, Y.; and Zheng, J. 2024. Prompt-based Generation of Natural Language Explanations of Synthetic Lethality for Cancer Drug Discovery. In Calzolari, N.; Kan, M.-Y.; Hoste, V.; Lenci, A.; Sakti, S.; and Xue, N., eds., Proceedings of the 2024 Joint International Con...

    2024

  51. [59]

    Zhang, X.; Liu, Q.; and Han, R. 2025. MMGCF: generating counterfactual explanations for molecular property prediction via motif rebuild. Journal of Computer and Communications, 13(1): 152--168

    2025

  52. [60]

    Zhao, Y.; Zhang, R.; Li, W.; Huang, D.; Guo, J.; Peng, S.; Hao, Y.; Wen, Y.; Hu, X.; Du, Z.; et al. 2024. Assessing and understanding creativity in large language models. arXiv preprint arXiv:2401.12491

    2024 arXiv

  53. [61]

    Y.; Yang, M.; Li, L.; May, L

    Zheng, Y.; Koh, H. Y.; Yang, M.; Li, L.; May, L. T.; Webb, G. I.; Pan, S.; and Church, G. 2024. Large language models in drug discovery and development: From disease mechanisms to clinical trials. arXiv preprint arXiv:2409.04481

    2024 arXiv

  54. [62]

    Zhong, Z.; Zhou, K.; and Mottin, D. 2024. Benchmarking large language models for molecule prediction tasks. arXiv preprint arXiv:2403.05075

    2024 arXiv

Pith tools

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