# PharmAgent: Constraint-Aware Search with Frozen Language Models for Molecular Optimization

arXiv: https://arxiv.org/abs/2610.05431 · HTML: https://mlsanity.com/paper/2610.05431/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 14 of its 33 references that are in the map.
- First posted on arXiv 4 Oct 2026.

## Builds on (most cited first, 14 of 14)

- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [Reflexion: Language Agents with Verbal Reinforcement Learning](https://mlsanity.com/paper/2303.11366/) (2023)
- [Convex optimization](https://mlsanity.com/paper/2106.01946/) (2021)
- [Billion-scale similarity search with GPUs](https://mlsanity.com/paper/1702.08734/) (2017)
- [Large Language Models as Optimizers](https://mlsanity.com/paper/2309.03409/) (2023)
- [Automatic chemical design using a data-driven continuous representation of molecules](https://mlsanity.com/paper/1610.02415/) (2016)
- [ChemCrow: Augmenting large-language models with chemistry tools](https://mlsanity.com/paper/2304.05376/) (2023)
- [The Synthesizability of Molecules Proposed by Generative Models](https://mlsanity.com/paper/2002.07007/) (2020)
- [Contextual Experience Replay for Self-Improvement of Language Agents](https://mlsanity.com/paper/2506.06698/) (2025)
- [Domain-Agnostic Molecular Generation with Chemical Feedback](https://mlsanity.com/paper/2301.11259/) (2023)
- [ExLLM: Experience-Enhanced LLM Optimization for Molecular Design and Beyond](https://mlsanity.com/paper/2502.12845/) (2025)
- [De Novo Molecular Design Enabled by Direct Preference Optimization and Curriculum Learning](https://mlsanity.com/paper/2504.01389/) (2025)
- [Uncertainty-Aware Multi-Objective Reinforcement Learning-Guided Diffusion Models for 3D De Novo Molecular Design](https://mlsanity.com/paper/2510.21153/) (2025)
- [InversionGNN: A Dual Path Network for Multi-Property Molecular Optimization](https://mlsanity.com/paper/2503.01488/) (2025)

## Abstract (arXiv)

Molecular optimization must improve target activity and satisfy developability constraints within limited evaluation budgets. Classical methods require tailored rules or training to incorporate chemical instructions and property feedback. Frozen language models can condition edits on this information, but need explicit constraint control and relevant experience. We therefore present PharmAgent, a constraint-aware molecular search method driven by adaptive external state. Its Lagrangian controller translates violations in accepted states into accumulated constraint pressure, keeping this history separate from current property measurements. Structure-indexed replay complements this feedback with relevant evaluated transitions that guide subsequent proposals. As a curriculum progressively activates constraints, candidates and the incumbent are compared under the same current objective, and the accepted state determines the next multiplier update. We derive an exact identity that characterizes how accepted-state violations accumulate in the controller's multipliers. Across five tasks with five independent runs, PharmAgent achieves a summed area under the target-score curves (AUC) of 3.9208 in target-only search, improving over MOLLEO by 37.3%. With online constraints, it achieves a property-adjusted AUC of 0.7076, improving over the strongest online baseline, ExLLM, by 53.8%. These results rank first among all evaluated methods in both target-only and constraint-aware search. The online comparison covers all five baseline frameworks. The full system leads every ablation variant in target quality, property-adjusted performance, and Pareto hypervolume. All five molecular cases reach feasible final states, documenting target gains and trade-offs.