---
title: "Auditing Discovery Claims — Chen 2026"
description: "The authors propose a two-sided audit. 62 references."
authors: ["Wenhui Chen", "Jianlin Chen", "Ziyao Lin", "Chi Man Vong"]
year: 2026
venue: "arXiv:2608.00981"
source: "https://arxiv.org/abs/2608.00981"
doi: "10.48550/arxiv.2608.00981"
arxiv: "2608.00981"
licence: "CC BY 4.0 (arXiv)"
licence_url: "http://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/chen-2026-auditing-discovery-claims-two/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/chen-2026-auditing-discovery-claims-two/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/4/) / Auditing Discovery Claims: A Two-Sided Criterion for Agentic Science, with the Negative Side Decidable

# Auditing Discovery Claims: A Two-Sided Criterion for Agentic Science, with the Negative Side Decidable

**Authors:** Wenhui Chen, Jianlin Chen, Ziyao Lin and Chi Man Vong · **Year:** 2026 · **Venue:** arXiv:2608.00981 · **Licence:** [CC BY 4.0 (arXiv)](http://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://arxiv.org/abs/2608.00981) · [PDF](https://arxiv.org/pdf/2608.00981) · [DOI 10.48550/arxiv.2608.00981](https://doi.org/10.48550/arxiv.2608.00981) · [arXiv 2608.00981](https://arxiv.org/abs/2608.00981) · [OpenAlex](https://openalex.org/W7172421479)

**Citations:** 0 citations (OpenAlex, 2026-10-14)

## Summary
Self-improving AI-for-science agents often back capability claims with benchmark gains, MDL gates or p-values. Such numbers cannot tell genuine progress from extra search, a shifted verifier, or exploitation of one imperfect oracle. The authors propose a two-sided audit. Its negative side is provable: a pseudoknot-free folding oracle (ViennaRNA) cannot express crossing base pairs, so what the agent's earlier self could certify is bounded offline. Its positive side is empirical, judged by a panel of crossing-aware predictors (pKiss, ProbKnot, ShapeKnots).

On RNA pseudoknot inverse design, an invented operator looks strong under the single predictor it optimizes but almost disappears under the full panel. Operators written by frontier LLMs, by contrast, beat a hand-built one under a predictor outside every loop while using less compute.

The work matters because it shows how one fallible oracle can inflate discovery claims. The authors also apply the audit to their own system and find it largely unimpressive. They concede that the panel's shared thermodynamic basis limits how far any conclusion reaches.

## Key points
- The single-oracle operator solves 43/60 crossing targets under pKiss but only 1/60 when all three predictors must agree. The context-free floor is 0/60.
- On the same 43 targets, the held-out ShapeKnots predictor confirms only 2 of the operator's designs, versus 26 for an MFE solver (p=8×10⁻⁷).
- Six frontier models (e.g., Claude-Opus-4.8, Gemini-3.1-pro, GLM-5.2) wrote operators from primitives. The two that ran without timeouts carried over at 0.293 versus 0.095 for the hand-built operator (n=951, p=5×10⁻⁵), with 4.6–10× fewer oracle calls.
- Of the three evidence rungs, the first two (adjudicated difference, not compute-driven) were reached. The third (a transferable mechanism) was not; seven candidates were tested.
- Pseudobase++ yields only 24 disjoint clusters, so a 100-target test set was generated from a formally defined crossing family. Two panel predictors agree at κ=0.673, which limits their independence.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

Fields: [AI & machine learning](https://md.research.elamur.ai/free/en/shelves/ai/)

## Institutions

- [South China University of Technology](https://md.research.elamur.ai/free/en/universities/south-china-university-of-technology/) — University, Guangzhou, China · 6 works · 1,323 citations
- [University of Macau](https://md.research.elamur.ai/free/en/universities/university-of-macau/) — University, Macao, Macao · 6 works · 1,192 citations

## Cite

Cite as: Wenhui Chen, Jianlin Chen, Ziyao Lin and Chi Man Vong. 2026. Auditing Discovery Claims: A Two-Sided Criterion for Agentic Science, with the Negative Side Decidable. arXiv:2608.00981.

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@misc{chen2026auditing,
  title = {{Auditing Discovery Claims: A Two-Sided Criterion for Agentic Science, with the Negative Side Decidable}},
  author = {Wenhui Chen, Jianlin Chen, Ziyao Lin and Chi Man Vong},
  year = {2026},
  doi = {10.48550/arxiv.2608.00981},
  eprint = {2608.00981},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2608.00981}
}
```

- [← previous Audio description between MT translation and recreation: An Interview Study for the Language Pair English-German](https://md.research.elamur.ai/free/en/papers/sauter-2026-audio-description-between-mt-translation/)
- [next → Augmenting LLM Reasoning with Dynamic Notes Writing for Complex MultiHop QA](https://md.research.elamur.ai/free/en/papers/maheshwary-2026-augmenting-llm-reasoning-dynamic-notes/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/4/)

## References

62 references, 7 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Jacobo Aguirre, Javier M. Buldú, Michael Stich, and Susanna C. Manrubia. Topological structure of the space of phenotypes: The case of RNA neutral networks. PLoS ONE, 6(10):e26324, 2011. doi: 10.1371/journal.pone.0026324. — [doi:10.1371/journal.pone.0026324](https://doi.org/10.1371/journal.pone.0026324) [find](https://scholar.google.com/scholar?q=Jacobo+Aguirre%2C+Javier+M%2E+Buld%C3%BA%2C+Michael+Stich%2C+and+Susanna+C%2E+Manrubia%2E+Topological+structure+of+the+space+of+phenotypes%3A+The+case+of+RNA+neutral+networks%2E+PLoS+ONE%2C+6%2810%29%3Ae26324%2C+2011%2E+doi%3A+10%2E1371%2F)
2. Dario Amodei, Chris Olah, Jacob Steinhardt, Paul Christiano, John Schulman, and Dan Mané. Concrete problems in AI safety. arXiv preprint arXiv:1606.06565, 2016. [our card](https://md.research.elamur.ai/free/en/papers/amodei-2016-concrete-problems-ai-safety/)
3. Ivan Anishchenko, Samuel J. Pellock, Tamuka M. Chidyausiku, et al. De novo protein design by deep network hallucination. Nature, 600:547–552, 2021. [find](https://scholar.google.com/scholar?q=Ivan+Anishchenko%2C+Samuel+J%2E+Pellock%2C+Tamuka+M%2E+Chidyausiku%2C+et+al%2E+De+novo+protein+design+by+deep+network+hallucination%2E+Nature%2C+600%3A547%E2%80%93552%2C+2021%2E)
4. Kirby A. Baker and Alden F. Pixley. Polynomial interpolation and the Chinese remainder theorem for algebraic systems. Mathematische Zeitschrift, 143(2):165–174, 1975. doi: 10.1007/ BF01187059. [find](https://scholar.google.com/scholar?q=Kirby+A%2E+Baker+and+Alden+F%2E+Pixley%2E+Polynomial+interpolation+and+the+Chinese+remainder+theorem+for+algebraic+systems%2E+Mathematische+Zeitschrift%2C+143%282%29%3A165%E2%80%93174%2C+1975%2E+doi%3A+10%2E1007%2F+BF01187059%2E)
5. Stanislav Bellaousov and David H. Mathews. ProbKnot: Fast prediction of RNA secondary structure including pseudoknots. RNA, 16(10):1870–1880, 2010. doi: 10.1261/rna.2125310. — [doi:10.1261/rna.2125310](https://doi.org/10.1261/rna.2125310) [find](https://scholar.google.com/scholar?q=Stanislav+Bellaousov+and+David+H%2E+Mathews%2E+ProbKnot%3A+Fast+prediction+of+RNA+secondary+structure+including+pseudoknots%2E+RNA%2C+16%2810%29%3A1870%E2%80%931880%2C+2010%2E+doi%3A+10%2E1261%2Frna%2E2125310%2E+%E2%80%94+%5Bdoi%3A10%2E1261%2Frna%2E2125310)
6. Édouard Bonnet, Paweł Rza̧żewski, and Florian Sikora. Designing RNA secondary structures is hard. Journal of Computational Biology, 27(3):302–316, 2020. doi: 10.1089/cmb.2019.0420. prelim. RECOMB 2018; arXiv:1710.11513. — [doi:10.1089/cmb.2019.0420](https://doi.org/10.1089/cmb.2019.0420) [find](https://scholar.google.com/scholar?q=%C3%89douard+Bonnet%2C+Pawe%C5%82+Rza%CC%A7%C5%BCewski%2C+and+Florian+Sikora%2E+Designing+RNA+secondary+structures+is+hard%2E+Journal+of+Computational+Biology%2C+27%283%29%3A302%E2%80%93316%2C+2020%2E+doi%3A+10%2E1089%2Fcmb%2E2019%2E0420%2E+prelim%2E+RECOMB+2018)
7. Anke Busch and Rolf Backofen. INFO-RNA—a fast approach to inverse RNA folding. Bioinformatics, 22(15):1823–1831, 2006. doi: 10.1093/bioinformatics/btl194. — [doi:10.1093/bioinformatics/btl194](https://doi.org/10.1093/bioinformatics/btl194) [find](https://scholar.google.com/scholar?q=Anke+Busch+and+Rolf+Backofen%2E+INFO%2DRNA%E2%80%94a+fast+approach+to+inverse+RNA+folding%2E+Bioinformatics%2C+22%2815%29%3A1823%E2%80%931831%2C+2006%2E+doi%3A+10%2E1093%2Fbioinformatics%2Fbtl194%2E+%E2%80%94+%5Bdoi%3A10%2E1093%2Fbioinformatics%2Fbtl194%5D%28https%3A%2F)
8. Wenhui Chen, Qingqing Mao, and Chenghua Wang. “solved” is a choice of verifier, not a fact: Deterministic cost–robustness frontiers of verifier choice in RNA and SAT. Technical report, InceptLabs-Shanghai, 2026. companion technical report. [find](https://scholar.google.com/scholar?q=Wenhui+Chen%2C+Qingqing+Mao%2C+and+Chenghua+Wang%2E+%E2%80%9Csolved%E2%80%9D+is+a+choice+of+verifier%2C+not+a+fact%3A+Deterministic+cost%E2%80%93robustness+frontiers+of+verifier+choice+in+RNA+and+SAT%2E+Technical+report%2C+InceptLabs%2DShan)
9. Noam Chomsky. Three models for the description of language. IRE Transactions on Information Theory, 2(3):113–124, 1956. doi: 10.1109/TIT.1956.1056813. [find](https://scholar.google.com/scholar?q=Noam+Chomsky%2E+Three+models+for+the+description+of+language%2E+IRE+Transactions+on+Information+Theory%2C+2%283%29%3A113%E2%80%93124%2C+1956%2E+doi%3A+10%2E1109%2FTIT%2E1956%2E1056813%2E)
10. Grégoire Delétang, Anian Ruoss, Jordi Grau-Moya, Tim Genewein, Li Kevin Wenliang, Elliot Catt, Chris Cundy, Marcus Hutter, Shane Legg, Joel Veness, and Pedro A. Ortega. Neural networks and the Chomsky hierarchy. In ICLR, 2023. arXiv:2207.02098. [our card](https://md.research.elamur.ai/free/en/papers/deletang-2022-neural-networks-chomsky-hierarchy/)
11. Robin D. Dowell and Sean R. Eddy. Evaluation of several lightweight stochastic context-free grammars for RNA secondary structure prediction. BMC Bioinformatics, 5:71, 2004. doi: 10.1186/1471-2105-5-71. — [doi:10.1186/1471-2105-5-71](https://doi.org/10.1186/1471-2105-5-71) [find](https://scholar.google.com/scholar?q=Robin+D%2E+Dowell+and+Sean+R%2E+Eddy%2E+Evaluation+of+several+lightweight+stochastic+context%2Dfree+grammars+for+RNA+secondary+structure+prediction%2E+BMC+Bioinformatics%2C+5%3A71%2C+2004%2E+doi%3A+10%2E1186%2F1471%2D2105%2D5%2D71)
12. Richard Durbin, Sean R. Eddy, Anders Krogh, and Graeme Mitchison. Biological Sequence Analysis. Cambridge University Press, 1998. [find](https://scholar.google.com/scholar?q=Richard+Durbin%2C+Sean+R%2E+Eddy%2C+Anders+Krogh%2C+and+Graeme+Mitchison%2E+Biological+Sequence+Analysis%2E+Cambridge+University+Press%2C+1998%2E)
13. Alhussein Fawzi, Matej Balog, Aja Huang, Thomas Hubert, Bernardino Romera-Paredes, Mohammadamin Barekatain, Alexander Novikov, Francisco J. R. Ruiz, Julian Schrittwieser, Grzegorz Swirszcz, David Silver, Demis Hassabis, and Pushmeet Kohli. Discovering faster matrix multiplication algorithms with rei… [find](https://scholar.google.com/scholar?q=Alhussein+Fawzi%2C+Matej+Balog%2C+Aja+Huang%2C+Thomas+Hubert%2C+Bernardino+Romera%2DParedes%2C+Mohammadamin+Barekatain%2C+Alexander+Novikov%2C+Francisco+J%2E+R%2E+Ruiz%2C+Julian+Schrittwieser%2C+Grzegorz+Swirszcz%2C+David+Silv)
14. Leo Gao, John Schulman, and Jacob Hilton. Scaling laws for reward model overoptimization. In ICML, PMLR 202, pages 10835–10866, 2023. arXiv:2210.10760. [our card](https://md.research.elamur.ai/free/en/papers/gao-2022-scaling-laws-reward-model-overoptimization/)
15. Juraj Gottweis, Wei-Hung Weng, Alexander Daryin, et al. Towards an AI co-scientist, 2025. arXiv:2502.18864. [our card](https://md.research.elamur.ai/free/en/papers/gottweis-2025-accelerating-scientific-discovery-co/)
16. Peter D. Grünwald. The Minimum Description Length Principle. MIT Press, 2007. [find](https://scholar.google.com/scholar?q=Peter+D%2E+Gr%C3%BCnwald%2E+The+Minimum+Description+Length+Principle%2E+MIT+Press%2C+2007%2E)
17. Christine E. Hajdin, Stanislav Bellaousov, Wayne Huggins, Christopher W. Leonard, David H. Mathews, and Kevin M. Weeks. Accurate SHAPE-directed RNA secondary structure modeling, including pseudoknots. PNAS, 110(14):5498–5503, 2013. doi: 10.1073/pnas.1219988110. — [doi:10.1073/pnas.1219988110](https://doi.org/10.1073/pnas.1219988110) [find](https://scholar.google.com/scholar?q=Christine+E%2E+Hajdin%2C+Stanislav+Bellaousov%2C+Wayne+Huggins%2C+Christopher+W%2E+Leonard%2C+David+H%2E+Mathews%2C+and+Kevin+M%2E+Weeks%2E+Accurate+SHAPE%2Ddirected+RNA+secondary+structure+modeling%2C+including+pseudoknots%2E)
18. John E. Hopcroft and Jeffrey D. Ullman. Introduction to Automata Theory, Languages, and Computation. Addison-Wesley, 1979. [find](https://scholar.google.com/scholar?q=John+E%2E+Hopcroft+and+Jeffrey+D%2E+Ullman%2E+Introduction+to+Automata+Theory%2C+Languages%2C+and+Computation%2E+Addison%2DWesley%2C+1979%2E)
19. Kexin Huang, Ying Jin, Ryan Li, Michael Y. Li, Emmanuel Candès, and Jure Leskovec. Automated hypothesis validation with agentic sequential falsifications. arXiv preprint arXiv:2502.09858, 2025. ICML 2025. [find](https://scholar.google.com/scholar?q=Kexin+Huang%2C+Ying+Jin%2C+Ryan+Li%2C+Michael+Y%2E+Li%2C+Emmanuel+Cand%C3%A8s%2C+and+Jure+Leskovec%2E+Automated+hypothesis+validation+with+agentic+sequential+falsifications%2E+arXiv+preprint+arXiv%3A2502%2E09858%2C+2025%2E+ICML+2)
20. Stefan Janssen and Robert Giegerich. The RNA shapes studio. Bioinformatics, 31(3):423–425, 2015. doi: 10.1093/bioinformatics/btu649. — [doi:10.1093/bioinformatics/btu649](https://doi.org/10.1093/bioinformatics/btu649) [find](https://scholar.google.com/scholar?q=Stefan+Janssen+and+Robert+Giegerich%2E+The+RNA+shapes+studio%2E+Bioinformatics%2C+31%283%29%3A423%E2%80%93425%2C+2015%2E+doi%3A+10%2E1093%2Fbioinformatics%2Fbtu649%2E+%E2%80%94+%5Bdoi%3A10%2E1093%2Fbioinformatics%2Fbtu649%5D%28https%3A%2F%2Fdoi%2Eorg%2F10%2E1093%2Fbioin)
21. Aravind K. Joshi. Tree adjoining grammars: How much context-sensitivity is required to provide reasonable structural descriptions? In David R. Dowty, Lauri Karttunen, and Arnold M. Zwicky, editors, Natural Language Parsing, pages 206–250. Cambridge University Press, 1985. [find](https://scholar.google.com/scholar?q=Aravind+K%2E+Joshi%2E+Tree+adjoining+grammars%3A+How+much+context%2Dsensitivity+is+required+to+provide+reasonable+structural+descriptions%3F+In+David+R%2E+Dowty%2C+Lauri+Karttunen%2C+and+Arnold+M%2E+Zwicky%2C+editors%2C+Na)
22. Yuki Kato, Hiroyuki Seki, and Tadao Kasami. RNA pseudoknotted structure prediction using stochastic multiple context-free grammar. IPSJ Digital Courier, 2:655–664, 2006. doi: 10. 2197/ipsjdc.2.655. [find](https://scholar.google.com/scholar?q=Yuki+Kato%2C+Hiroyuki+Seki%2C+and+Tadao+Kasami%2E+RNA+pseudoknotted+structure+prediction+using+stochastic+multiple+context%2Dfree+grammar%2E+IPSJ+Digital+Courier%2C+2%3A655%E2%80%93664%2C+2006%2E+doi%3A+10%2E+2197%2Fipsjdc%2E2%2E655%2E)
23. Bjarne Knudsen and Jotun Hein. RNA secondary structure prediction using stochastic context-free grammars and evolutionary history. Bioinformatics, 15(6):446–454, 1999. doi: 10.1093/ bioinformatics/15.6.446. [find](https://scholar.google.com/scholar?q=Bjarne+Knudsen+and+Jotun+Hein%2E+RNA+secondary+structure+prediction+using+stochastic+context%2Dfree+grammars+and+evolutionary+history%2E+Bioinformatics%2C+15%286%29%3A446%E2%80%93454%2C+1999%2E+doi%3A+10%2E1093%2F+bioinformatics%2F15%2E)
24. Rohan V. Koodli, Boris Rudolfs, Hannah K. Wayment-Steele, and Rhiju Das. Redesigning the Eterna100 for the Vienna 2 folding engine. bioRxiv, 2021. doi: 10.1101/2021.08.26.457839. preprint. — [doi:10.1101/2021.08.26.457839](https://doi.org/10.1101/2021.08.26.457839) [find](https://scholar.google.com/scholar?q=Rohan+V%2E+Koodli%2C+Boris+Rudolfs%2C+Hannah+K%2E+Wayment%2DSteele%2C+and+Rhiju+Das%2E+Redesigning+the+Eterna100+for+the+Vienna+2+folding+engine%2E+bioRxiv%2C+2021%2E+doi%3A+10%2E1101%2F2021%2E08%2E26%2E457839%2E+preprint%2E+%E2%80%94+%5Bdoi%3A10%2E1)
25. Roham Koohestani, Ziyou Li, Anton Podkopaev, and Maliheh Izadi. Are agents just automata? on the formal equivalence between agentic AI and the Chomsky hierarchy. arXiv preprint arXiv:2510.23487, 2025. — [arXiv:2510.23487](https://arxiv.org/abs/2510.23487) [find](https://scholar.google.com/scholar?q=Roham+Koohestani%2C+Ziyou+Li%2C+Anton+Podkopaev%2C+and+Maliheh+Izadi%2E+Are+agents+just+automata%3F+on+the+formal+equivalence+between+agentic+AI+and+the+Chomsky+hierarchy%2E+arXiv+preprint+arXiv%3A2510%2E23487%2C+2025%2E)
26. Imre Lakatos. The Methodology of Scientific Research Programmes. Cambridge University Press, 1978. [find](https://scholar.google.com/scholar?q=Imre+Lakatos%2E+The+Methodology+of+Scientific+Research+Programmes%2E+Cambridge+University+Press%2C+1978%2E)
27. Hunter Lightman, Vineet Kosaraju, Yura Burda, Harri Edwards, Bowen Baker, Teddy Lee, Jan Leike, John Schulman, Ilya Sutskever, and Karl Cobbe. Let’s verify step by step. arXiv preprint arXiv:2305.20050, 2023. ICLR 2024. [our card](https://md.research.elamur.ai/free/en/papers/lightman-2023-let-s-verify-step-step/)
28. Zijun Liu, Kaiming Liu, Yiqi Zhu, Xuanyu Lei, Zonghan Yang, Zhenhe Zhang, Peng Li, and Yang Liu. AIGS: Generating science from AI-powered automated falsification. arXiv preprint arXiv:2411.11910, 2024. — [arXiv:2411.11910](https://arxiv.org/abs/2411.11910) [find](https://scholar.google.com/scholar?q=Zijun+Liu%2C+Kaiming+Liu%2C+Yiqi+Zhu%2C+Xuanyu+Lei%2C+Zonghan+Yang%2C+Zhenhe+Zhang%2C+Peng+Li%2C+and+Yang+Liu%2E+AIGS%3A+Generating+science+from+AI%2Dpowered+automated+falsification%2E+arXiv+preprint+arXiv%3A2411%2E11910%2C+2024)
29. Ronny Lorenz, Stephan H. Bernhart, Christian Höner zu Siederdissen, Hakim Tafer, Christoph Flamm, Peter F. Stadler, and Ivo L. Hofacker. ViennaRNA package 2.0. Algorithms for Molecular Biology, 6(1):26, 2011. doi: 10.1186/1748-7188-6-26. — [doi:10.1186/1748-7188-6-26](https://doi.org/10.1186/1748-7188-6-26) [find](https://scholar.google.com/scholar?q=Ronny+Lorenz%2C+Stephan+H%2E+Bernhart%2C+Christian+H%C3%B6ner+zu+Siederdissen%2C+Hakim+Tafer%2C+Christoph+Flamm%2C+Peter+F%2E+Stadler%2C+and+Ivo+L%2E+Hofacker%2E+ViennaRNA+package+2%2E0%2E+Algorithms+for+Molecular+Biology%2C+6%281%29%3A2)
30. Rune B. Lyngsø and Christian N. S. Pedersen. RNA pseudoknot prediction in energy-based models. Journal of Computational Biology, 7(3-4):409–427, 2000. doi: 10.1089/ 106652700750050862. [find](https://scholar.google.com/scholar?q=Rune+B%2E+Lyngs%C3%B8+and+Christian+N%2E+S%2E+Pedersen%2E+RNA+pseudoknot+prediction+in+energy%2Dbased+models%2E+Journal+of+Computational+Biology%2C+7%283%2D4%29%3A409%E2%80%93427%2C+2000%2E+doi%3A+10%2E1089%2F+106652700750050862%2E)
31. Daniel J. Mankowitz, Andrea Michi, Anton Zhernov, Marco Gelmi, Marco Selvi, Cosmin Paduraru, et al. Faster sorting algorithms discovered using deep reinforcement learning. Nature, 618 (7964):257–263, 2023. doi: 10.1038/s41586-023-06004-9. [find](https://scholar.google.com/scholar?q=Daniel+J%2E+Mankowitz%2C+Andrea+Michi%2C+Anton+Zhernov%2C+Marco+Gelmi%2C+Marco+Selvi%2C+Cosmin+Paduraru%2C+et+al%2E+Faster+sorting+algorithms+discovered+using+deep+reinforcement+learning%2E+Nature%2C+618+%287964%29%3A257%E2%80%93263%2C+)
32. John S. McCaskill. The equilibrium partition function and base pair binding probabilities for RNA secondary structure. Biopolymers, 29(6-7):1105–1119, 1990. doi: 10.1002/bip.360290621. — [doi:10.1002/bip.360290621](https://doi.org/10.1002/bip.360290621) [find](https://scholar.google.com/scholar?q=John+S%2E+McCaskill%2E+The+equilibrium+partition+function+and+base+pair+binding+probabilities+for+RNA+secondary+structure%2E+Biopolymers%2C+29%286%2D7%29%3A1105%E2%80%931119%2C+1990%2E+doi%3A+10%2E1002%2Fbip%2E360290621%2E+%E2%80%94+%5Bdoi%3A10%2E1002%2F)
33. Markus E. Nebel and Frank Weinberg. Algebraic and combinatorial properties of common RNA pseudoknot classes. Journal of Computational Biology, 19(10):1134–1150, 2012. PMC3469209. [find](https://scholar.google.com/scholar?q=Markus+E%2E+Nebel+and+Frank+Weinberg%2E+Algebraic+and+combinatorial+properties+of+common+RNA+pseudoknot+classes%2E+Journal+of+Computational+Biology%2C+19%2810%29%3A1134%E2%80%931150%2C+2012%2E+PMC3469209%2E)
34. Alexander Novikov, Ngân Vũ, Marvin Eisenberger, Emilien Dupont, Po-Sen Huang, Adam Zsolt Wagner, Sergey Shirobokov, et al. AlphaEvolve: A coding agent for scientific and algorithmic discovery. arXiv preprint arXiv:2506.13131, 2025. [our card](https://md.research.elamur.ai/free/en/papers/novikov-2025-alphaevolve-coding-agent-scientific-algorithmic/)
35. Ruth Nussinov and Ann B. Jacobson. Fast algorithm for predicting the secondary structure of single-stranded RNA. PNAS, 77(11):6309–6313, 1980. doi: 10.1073/pnas.77.11.6309. — [doi:10.1073/pnas.77.11.6309](https://doi.org/10.1073/pnas.77.11.6309) [find](https://scholar.google.com/scholar?q=Ruth+Nussinov+and+Ann+B%2E+Jacobson%2E+Fast+algorithm+for+predicting+the+secondary+structure+of+single%2Dstranded+RNA%2E+PNAS%2C+77%2811%29%3A6309%E2%80%936313%2C+1980%2E+doi%3A+10%2E1073%2Fpnas%2E77%2E11%2E6309%2E+%E2%80%94+%5Bdoi%3A10%2E1073%2Fpnas%2E77%2E11%2E6)
36. William Ogden. A helpful result for proving inherent ambiguity. Mathematical Systems Theory, 2(3):191–194, 1968. doi: 10.1007/BF01694004. — [doi:10.1007/bf01694004](https://doi.org/10.1007/bf01694004) [find](https://scholar.google.com/scholar?q=William+Ogden%2E+A+helpful+result+for+proving+inherent+ambiguity%2E+Mathematical+Systems+Theory%2C+2%283%29%3A191%E2%80%93194%2C+1968%2E+doi%3A+10%2E1007%2FBF01694004%2E+%E2%80%94+%5Bdoi%3A10%2E1007%2Fbf01694004%5D%28https%3A%2F%2Fdoi%2Eorg%2F10%2E1007%2Fbf01694004%29)
37. Karl R. Popper. The Logic of Scientific Discovery. Hutchinson, London, 1959. transl. of Logik der Forschung, 1934. [find](https://scholar.google.com/scholar?q=Karl+R%2E+Popper%2E+The+Logic+of+Scientific+Discovery%2E+Hutchinson%2C+London%2C+1959%2E+transl%2E+of+Logik+der+Forschung%2C+1934%2E)
38. Jens Reeder and Robert Giegerich. Design, implementation and evaluation of a practical pseudoknot folding algorithm based on thermodynamics. BMC Bioinformatics, 5:104, 2004. doi: 10.1186/1471-2105-5-104. — [doi:10.1186/1471-2105-5-104](https://doi.org/10.1186/1471-2105-5-104) [find](https://scholar.google.com/scholar?q=Jens+Reeder+and+Robert+Giegerich%2E+Design%2C+implementation+and+evaluation+of+a+practical+pseudoknot+folding+algorithm+based+on+thermodynamics%2E+BMC+Bioinformatics%2C+5%3A104%2C+2004%2E+doi%3A+10%2E1186%2F1471%2D2105%2D5%2D1)
39. Elena Rivas and Sean R. Eddy. A dynamic programming algorithm for RNA structure prediction including pseudoknots. Journal of Molecular Biology, 285(5):2053–2068, 1999. doi: 10.1006/jmbi.1998.2436. — [doi:10.1006/jmbi.1998.2436](https://doi.org/10.1006/jmbi.1998.2436) [find](https://scholar.google.com/scholar?q=Elena+Rivas+and+Sean+R%2E+Eddy%2E+A+dynamic+programming+algorithm+for+RNA+structure+prediction+including+pseudoknots%2E+Journal+of+Molecular+Biology%2C+285%285%29%3A2053%E2%80%932068%2C+1999%2E+doi%3A+10%2E1006%2Fjmbi%2E1998%2E2436%2E+%E2%80%94+%5B)
40. Elena Rivas and Sean R. Eddy. The language of RNA: A formal grammar that includes pseudoknots. Bioinformatics, 16(4):334–340, 2000. doi: 10.1093/bioinformatics/16.4.334. — [doi:10.1093/bioinformatics/16.4.334](https://doi.org/10.1093/bioinformatics/16.4.334) [find](https://scholar.google.com/scholar?q=Elena+Rivas+and+Sean+R%2E+Eddy%2E+The+language+of+RNA%3A+A+formal+grammar+that+includes+pseudoknots%2E+Bioinformatics%2C+16%284%29%3A334%E2%80%93340%2C+2000%2E+doi%3A+10%2E1093%2Fbioinformatics%2F16%2E4%2E334%2E+%E2%80%94+%5Bdoi%3A10%2E1093%2Fbioinformatics%2F)
41. RNAInvBench. RNAInvBench and Pseudobase++: Pseudoknotted RNA inverse-design targets, 2024. inverse-design benchmark suite. [find](https://scholar.google.com/scholar?q=RNAInvBench%2E+RNAInvBench+and+Pseudobase%2B%2B%3A+Pseudoknotted+RNA+inverse%2Ddesign+targets%2C+2024%2E+inverse%2Ddesign+benchmark+suite%2E)
42. Bernardino Romera-Paredes, Mohammadamin Barekatain, Alexander Novikov, Matej Balog, M. Pawan Kumar, Emilien Dupont, Francisco J. R. Ruiz, Jordan S. Ellenberg, Pengming Wang, Omar Fawzi, Pushmeet Kohli, and Alhussein Fawzi. Mathematical discoveries from program search with large language models. Natu… [our card](https://md.research.elamur.ai/free/en/papers/paredes-2023-mathematical-discoveries-program-search-large/)
43. Frederic Runge, Danny Stoll, Stefan Falkner, and Frank Hutter. Learning to design RNA. In ICLR, 2019. arXiv:1812.11951. — [arXiv:1812.11951](https://arxiv.org/abs/1812.11951) [find](https://scholar.google.com/scholar?q=Frederic+Runge%2C+Danny+Stoll%2C+Stefan+Falkner%2C+and+Frank+Hutter%2E+Learning+to+design+RNA%2E+In+ICLR%2C+2019%2E+arXiv%3A1812%2E11951%2E+%E2%80%94+%5BarXiv%3A1812%2E11951%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F1812%2E11951%29)
44. Yasubumi Sakakibara, Michael Brown, Richard Hughey, I. Saira Mian, Kimmen Sjölander, Rebecca C. Underwood, and David Haussler. Stochastic context-free grammars for tRNA modeling. Nucleic Acids Research, 22(23):5112–5120, 1994. doi: 10.1093/nar/22.23.5112. — [doi:10.1093/nar/22.23.5112](https://doi.org/10.1093/nar/22.23.5112) [find](https://scholar.google.com/scholar?q=Yasubumi+Sakakibara%2C+Michael+Brown%2C+Richard+Hughey%2C+I%2E+Saira+Mian%2C+Kimmen+Sj%C3%B6lander%2C+Rebecca+C%2E+Underwood%2C+and+David+Haussler%2E+Stochastic+context%2Dfree+grammars+for+tRNA+modeling%2E+Nucleic+Acids+Researc)
45. Thomas J. Schaefer. The complexity of satisfiability problems. In STOC, pages 216–226, 1978. doi: 10.1145/800133.804350. — [doi:10.1145/800133.804350](https://doi.org/10.1145/800133.804350) [find](https://scholar.google.com/scholar?q=Thomas+J%2E+Schaefer%2E+The+complexity+of+satisfiability+problems%2E+In+STOC%2C+pages+216%E2%80%93226%2C+1978%2E+doi%3A+10%2E1145%2F800133%2E804350%2E+%E2%80%94+%5Bdoi%3A10%2E1145%2F800133%2E804350%5D%28https%3A%2F%2Fdoi%2Eorg%2F10%2E1145%2F800133%2E804350%29)
46. Peter Schuster, Walter Fontana, Peter F. Stadler, and Ivo L. Hofacker. From sequences to shapes and back: A case study in RNA secondary structures. Proceedings of the Royal Society of London B, 255(1344):279–284, 1994. doi: 10.1098/rspb.1994.0040. — [doi:10.1098/rspb.1994.0040](https://doi.org/10.1098/rspb.1994.0040) [find](https://scholar.google.com/scholar?q=Peter+Schuster%2C+Walter+Fontana%2C+Peter+F%2E+Stadler%2C+and+Ivo+L%2E+Hofacker%2E+From+sequences+to+shapes+and+back%3A+A+case+study+in+RNA+secondary+structures%2E+Proceedings+of+the+Royal+Society+of+London+B%2C+255%2813)
47. David B. Searls. The linguistics of DNA. American Scientist, 80(6):579–591, 1992. [find](https://scholar.google.com/scholar?q=David+B%2E+Searls%2E+The+linguistics+of+DNA%2E+American+Scientist%2C+80%286%29%3A579%E2%80%93591%2C+1992%2E)
48. David B. Searls. The language of genes. Nature, 420(6912):211–217, 2002. doi: 10.1038/ nature01255. [find](https://scholar.google.com/scholar?q=David+B%2E+Searls%2E+The+language+of+genes%2E+Nature%2C+420%286912%29%3A211%E2%80%93217%2C+2002%2E+doi%3A+10%2E1038%2F+nature01255%2E)
49. Hiroyuki Seki, Takashi Matsumura, Mamoru Fujii, and Tadao Kasami. On multiple context-free grammars. Theoretical Computer Science, 88(2):191–229, 1991. doi: 10.1016/0304-3975(91) 90374-B. — [doi:10.1016/0304-3975(91)](https://doi.org/10.1016/0304-3975%2891%29) [find](https://scholar.google.com/scholar?q=Hiroyuki+Seki%2C+Takashi+Matsumura%2C+Mamoru+Fujii%2C+and+Tadao+Kasami%2E+On+multiple+context%2Dfree+grammars%2E+Theoretical+Computer+Science%2C+88%282%29%3A191%E2%80%93229%2C+1991%2E+doi%3A+10%2E1016%2F0304%2D3975%2891%29+90374%2DB%2E+%E2%80%94+%5Bdoi%3A10%2E10)
50. Joar Skalse, Nikolaus H. R. Howe, Dmitrii Krasheninnikov, and David Krueger. Defining and characterizing reward hacking. In NeurIPS, volume 35, 2022. arXiv:2209.13085. — [arXiv:2209.13085](https://arxiv.org/abs/2209.13085) [find](https://scholar.google.com/scholar?q=Joar+Skalse%2C+Nikolaus+H%2E+R%2E+Howe%2C+Dmitrii+Krasheninnikov%2C+and+David+Krueger%2E+Defining+and+characterizing+reward+hacking%2E+In+NeurIPS%2C+volume+35%2C+2022%2E+arXiv%3A2209%2E13085%2E+%E2%80%94+%5BarXiv%3A2209%2E13085%5D%28https%3A%2F%2Farx)
51. K. Vijay-Shanker and David J. Weir. The equivalence of four extensions of context-free grammars. Mathematical Systems Theory, 27(6):511–546, 1994. doi: 10.1007/BF01191624. — [doi:10.1007/bf01191624](https://doi.org/10.1007/bf01191624) [find](https://scholar.google.com/scholar?q=K%2E+Vijay%2DShanker+and+David+J%2E+Weir%2E+The+equivalence+of+four+extensions+of+context%2Dfree+grammars%2E+Mathematical+Systems+Theory%2C+27%286%29%3A511%E2%80%93546%2C+1994%2E+doi%3A+10%2E1007%2FBF01191624%2E+%E2%80%94+%5Bdoi%3A10%2E1007%2Fbf01191624%5D%28h)
52. Fiona Y. Wang and Markus J. Buehler. Self-revising discovery systems for science: A categorical framework for agentic AI. arXiv preprint arXiv:2606.01444, 2026. — [arXiv:2606.01444](https://arxiv.org/abs/2606.01444) [find](https://scholar.google.com/scholar?q=Fiona+Y%2E+Wang+and+Markus+J%2E+Buehler%2E+Self%2Drevising+discovery+systems+for+science%3A+A+categorical+framework+for+agentic+AI%2E+arXiv+preprint+arXiv%3A2606%2E01444%2C+2026%2E+%E2%80%94+%5BarXiv%3A2606%2E01444%5D%28https%3A%2F%2Farxiv%2Eorg%2F)
53. Guanzhi Wang, Yuqi Xie, Yunfan Jiang, et al. Voyager: An open-ended embodied agent with large language models, 2023. arXiv:2305.16291. [find](https://scholar.google.com/scholar?q=Guanzhi+Wang%2C+Yuqi+Xie%2C+Yunfan+Jiang%2C+et+al%2E+Voyager%3A+An+open%2Dended+embodied+agent+with+large+language+models%2C+2023%2E+arXiv%3A2305%2E16291%2E)
54. Joseph L. Watson, David Juergens, Nathaniel R. Bennett, et al. De novo design of protein structure and function with RFdiffusion. Nature, 620:1089–1100, 2023. [find](https://scholar.google.com/scholar?q=Joseph+L%2E+Watson%2C+David+Juergens%2C+Nathaniel+R%2E+Bennett%2C+et+al%2E+De+novo+design+of+protein+structure+and+function+with+RFdiffusion%2E+Nature%2C+620%3A1089%E2%80%931100%2C+2023%2E)
55. Jiaqi Wei, Yuejin Yang, Xiang Zhang, et al. From AI for science to agentic science: A survey on autonomous scientific discovery, 2025. arXiv:2508.14111. [find](https://scholar.google.com/scholar?q=Jiaqi+Wei%2C+Yuejin+Yang%2C+Xiang+Zhang%2C+et+al%2E+From+AI+for+science+to+agentic+science%3A+A+survey+on+autonomous+scientific+discovery%2C+2025%2E+arXiv%3A2508%2E14111%2E)
56. Tomasz K. Wirecki, Grzegorz Lach, Nagendar Goud Badepally, S. Naeim Moafinejad, Farhang Jaryani, Gaja Klaudel, Kalina Nec, Eugene F. Baulin, and Janusz M. Bujnicki. DesiRNA: Structure-based design of RNA sequences with a replica exchange monte carlo approach. Nucleic Acids Research, 53(2), 2025. doi… — [doi:10.1093/nar/gkae1306](https://doi.org/10.1093/nar/gkae1306) [find](https://scholar.google.com/scholar?q=Tomasz+K%2E+Wirecki%2C+Grzegorz+Lach%2C+Nagendar+Goud+Badepally%2C+S%2E+Naeim+Moafinejad%2C+Farhang+Jaryani%2C+Gaja+Klaudel%2C+Kalina+Nec%2C+Eugene+F%2E+Baulin%2C+and+Janusz+M%2E+Bujnicki%2E+DesiRNA%3A+Structure%2Dbased+design+of+)
57. Ce Xu et al. Position: Falsify, don’t just discover—AI-generated discoveries are not born scientific. In ICML, Position Paper Track, 2025. OpenReview SlgXCLZFj3. [find](https://scholar.google.com/scholar?q=Ce+Xu+et+al%2E+Position%3A+Falsify%2C+don%E2%80%99t+just+discover%E2%80%94AI%2Dgenerated+discoveries+are+not+born+scientific%2E+In+ICML%2C+Position+Paper+Track%2C+2025%2E+OpenReview+SlgXCLZFj3%2E)
58. Joseph N. Zadeh, Conrad D. Steenberg, Justin S. Bois, Brian R. Wolfe, Marshall B. Pierce, Asif R. Khan, Robert M. Dirks, and Niles A. Pierce. NUPACK: Analysis and design of nucleic acid systems. Journal of Computational Chemistry, 32(1):170–173, 2011. doi: 10.1002/jcc. 21596. — [doi:10.1002/jcc](https://doi.org/10.1002/jcc) [find](https://scholar.google.com/scholar?q=Joseph+N%2E+Zadeh%2C+Conrad+D%2E+Steenberg%2C+Justin+S%2E+Bois%2C+Brian+R%2E+Wolfe%2C+Marshall+B%2E+Pierce%2C+Asif+R%2E+Khan%2C+Robert+M%2E+Dirks%2C+and+Niles+A%2E+Pierce%2E+NUPACK%3A+Analysis+and+design+of+nucleic+acid+systems%2E+Journ)
59. Tianshuo Zhou, Ning Dai, Sizhen Li, Max Ward, David H. Mathews, and Liang Huang. RNA design via structure-aware multifrontier ensemble optimization. Bioinformatics, 39(Supplement 1):i563–i571, 2023. doi: 10.1093/bioinformatics/btad252. — [doi:10.1093/bioinformatics/btad252](https://doi.org/10.1093/bioinformatics/btad252) [find](https://scholar.google.com/scholar?q=Tianshuo+Zhou%2C+Ning+Dai%2C+Sizhen+Li%2C+Max+Ward%2C+David+H%2E+Mathews%2C+and+Liang+Huang%2E+RNA+design+via+structure%2Daware+multifrontier+ensemble+optimization%2E+Bioinformatics%2C+39%28Supplement+1%29%3Ai563%E2%80%93i571%2C+2023%2E+d)
60. Tianshuo Zhou, Wei Yu Tang, David H. Mathews, and Liang Huang. Undesignable RNA structure identification via rival structure generation and structure decomposition. In RECOMB, LNCS 14758, pages 270–287, 2024. doi: 10.1007/978-1-0716-3989-4_17. arXiv:2311.08339. — [doi:10.1007/978-1-0716-3989-4_17](https://doi.org/10.1007/978-1-0716-3989-4_17) [find](https://scholar.google.com/scholar?q=Tianshuo+Zhou%2C+Wei+Yu+Tang%2C+David+H%2E+Mathews%2C+and+Liang+Huang%2E+Undesignable+RNA+structure+identification+via+rival+structure+generation+and+structure+decomposition%2E+In+RECOMB%2C+LNCS+14758%2C+pages+270%E2%80%9328)
61. Tianshuo Zhou, Apoorv Malik, Wei Yu Tang, David H. Mathews, and Liang Huang. Scalable and interpretable identification of minimal undesignable RNA structure motifs with rotational invariance. In RECOMB, 2025. arXiv:2402.17206. — [arXiv:2402.17206](https://arxiv.org/abs/2402.17206) [find](https://scholar.google.com/scholar?q=Tianshuo+Zhou%2C+Apoorv+Malik%2C+Wei+Yu+Tang%2C+David+H%2E+Mathews%2C+and+Liang+Huang%2E+Scalable+and+interpretable+identification+of+minimal+undesignable+RNA+structure+motifs+with+rotational+invariance%2E+In+RECOM)
62. Michael Zuker and Patrick Stiegler. Optimal computer folding of large RNA sequences using thermodynamics and auxiliary information. Nucleic Acids Research, 9(1):133–148, 1981. doi: 10.1093/nar/9.1.133. — [doi:10.1093/nar/9.1.133](https://doi.org/10.1093/nar/9.1.133) [find](https://scholar.google.com/scholar?q=Michael+Zuker+and+Patrick+Stiegler%2E+Optimal+computer+folding+of+large+RNA+sequences+using+thermodynamics+and+auxiliary+information%2E+Nucleic+Acids+Research%2C+9%281%29%3A133%E2%80%93148%2C+1981%2E+doi%3A+10%2E1093%2Fnar%2F9%2E1%2E133)