arXiv:2606. 02386v1 Announce Type: new Abstract: Protein language models (PLMs) are passive oracles: they generate sequences in a single forward pass with no mechanism to consult external biophysical feedback or redirect generation when a candidate violates thermodynamic or structural constraints.
arXiv:2602.00663v3 Announce Type: replace Abstract: Optimizing molecules to achieve desired properties is a central bottleneck across the chemical sciences, particularly in the pharmaceutical industr...
arXiv:2606. 18961v1 Announce Type: new Abstract: Protein language models (PLMs) have emerged as powerful tools for controllable biomolecular design, yet their post-training adaptation typically relies on costly wet-lab validation or curated preference datasets.
arXiv:2608. 12192v1 Announce Type: new Abstract: Foundation models for protein structure prediction remain unreliable on certain targets.
Oracle-limited molecular optimization gives reward only after a complete molecule is generated, while each rollout requires many local next-token decisions. This delayed-feedback interface makes molecular policy optimization myopic: an optimizer can learn that a molecule was good without knowing which intermediate actions made it good.
arXiv:2607. 26391v1 Announce Type: new Abstract: Oracle-limited molecular optimization gives reward only after a complete molecule is generated, while each rollout requires many local next-token decisions.
Protein language models (PLMs) have emerged as powerful tools for controllable biomolecular design, yet their post-training adaptation typically relies on costly wet-lab validation or curated preference datasets. To overcome this supervision bottleneck, we introduce unsupervised reward optimization of PLMs, a comprehensive framework for steerable protein generation without ground-truth labels.
arXiv:2606. 11256v1 Announce Type: cross Abstract: Designing molecules with target properties is most useful when candidate structures are accompanied by feasible synthetic routes.
AgentFold is a multi‑agent framework that treats protein‑folding model design as a closed‑loop search over executable code variants. Starting from the ESMFold codebase, the agents generate hypotheses, modify and debug code, evaluate model variants, and store both successes and failures in structured memory, guided by an MCTS‑style policy that allocates GPU resources. In an engineering‑scale experiment, AgentFold explored about 80 variants using 5,000 GPU‑hours and 170 million LLM tokens, improving the best lDDT score by 7.5% over independent Codex proposals and outperforming a random‑search baseline, while also uncovering empirical design patterns such as the benefits of early, soft, learnable priors.
Foundation models for protein structure prediction remain unreliable on certain targets. External oracles can flag and correct these failures, but biological oracles are expensive, making oracle budget a critical constraint.
The study evaluates 4‑bit quantization and low‑rank adapter fine‑tuning (QLoRA) on several large protein language models, finding that many model‑task pairs retain over 90% of full fine‑tuning performance while achieving up to 90% GPU memory savings. QLoRA preserves early‑layer representations and induces task‑specific changes in later layers, closely resembling full fine‑tuning with smaller representational shifts. For generative models, 4‑bit quantization largely maintains structural and sequence‑level properties, though token‑level analysis reveals model‑dependent changes in autoregressive output distributions.
arXiv:2604.07669v3 Announce Type: replace-cross Abstract: Synthesizable molecular optimization seeks to improve target properties while ensuring that molecular modifications follow feasible synthetic...
The paper introduces OmicsBench, a new reasoning benchmark for multi‑omics sequences that includes 1,160 expert‑validated questions across DNA regulation, RNA processing, and protein function tasks, requiring traceable evidence chains. Evaluation of 17 large language models shows that scientific LLMs, while more accurate in classification, often lack valid evidence, suggesting shortcut learning. To address this, the authors propose tool‑augmented on‑policy distillation (TA‑OPD), a post‑training method that improves both evidence grounding and predictive performance across five Qwen3.5 models of varying sizes.