arXiv:2605. 00182v3 Announce Type: replace Abstract: Proteins are shaped by gradual evolution under biophysical and functional constraints.
arXiv:2609.37675v1 Announce Type: new Abstract: Protein Language Models (PLMs) have made remarkable progress following scaling laws established in natural language processing across sequence- and str...
arXiv:2510.03095v4 Announce Type: replace Abstract: Diffusion- and flow-based generative models have recently demonstrated strong performance in protein backbone generation tasks, offering unpreceden...
arXiv:2603. 14717v2 Announce Type: replace Abstract: Generating novel protein sequences that respect a family's statistical constraints typically requires training deep generative models on thousands to millions of examples.
Spectral Feedback is a new algorithm for aligning discrete diffusion models at test time by iteratively revisiting and editing token positions rather than only steering the reverse process. It selects edit-sets—groups of token positions to re-mask and re-sample—using sparse Fourier representations of edit-set value functions, enabling efficient optimization of which tokens to revisit. The method is model-agnostic and improves alignment performance across pretrained, test‑time aligned, and fine‑tuned diffusion models, achieving significant gains in protein stability for inverse folding tasks.
arXiv:2512. 15133v3 Announce Type: replace-cross Abstract: Proteins inherently possess a consistent sequence-structure duality.
arXiv:2606. 10543v1 Announce Type: cross Abstract: Designing functional biological sequences requires navigating vast discrete spaces under strict evolutionary and biophysical constraints.
arXiv:2509. 26405v2 Announce Type: replace Abstract: We introduce InVirtuoGen, a discrete flow generative model for fragmented SMILES for de novo and fragment-constrained generation, and target-property/lead optimization of small molecules.
The paper introduces GenDA, a bidirectional discrete diffusion model designed for genomic sequence reconstruction, hypothesizing that entropy-guided span placement would improve variant-effect prediction and functional sequence generation. While the 202‑million‑parameter GenDA model achieves a higher ClinVar SNV AUROC (0.774) than a comparable autoregressive model, the improvement is not attributable to entropy guidance, and the model fails to outperform a shuffled‑gap baseline in zero‑shot functional inpainting across various genomic regions. The authors identify limitations such as tokenization granularity, span length caps, and the mismatch between local sequence complexity and functional importance, concluding that variant prediction, corruption priors, and functional generation are distinct tasks requiring separate validation.
ProtLingo is a protein language modeling framework that enhances a pretrained single‑sequence Transformer backbone with conditional local memory and sparse expert routing. It maps residue representations into discrete codes, composes local windows into latent N‑gram addresses, and retrieves reusable residual signals for recurring sequence contexts. The model also converts selected feed‑forward blocks into sparse Mixture‑of‑Experts layers, allowing residue‑dependent computation while activating only a subset of parameters, achieving competitive performance on protein fitness prediction, FLIP benchmarks, and supervised contact prediction with a 150M‑parameter backbone.
arXiv:2606. 02133v1 Announce Type: cross Abstract: Non-monotonic sequence generation methods, such as masked diffusion models, provide a flexible alternative to left-to-right autoregressive modeling by allowing tokens to be generated in non-fixed and prescribed orders.
arXiv:2606. 16044v1 Announce Type: new Abstract: Protein language models (pLMs) can generate novel protein sequences with properties beyond those observed in nature, yet the mechanisms underlying protein generation remain poorly understood.
arXiv:2507. 08920v4 Announce Type: replace-cross Abstract: We introduce AMix-1, a powerful protein foundation model built on Bayesian Flow Networks and empowered by a systematic training methodology, encompassing pretraining scaling laws, emergent capability analysis, in-context learning mechanism, and test-time scaling algorithm.