Skip to content

When Edit Flows are Edit Jumps: replicating Edit Flows and EvoFlows

Sep 2026 · arXiv.org · Vol abs/2609.18745 · 0 citations · 28 references
Computer Science Mathematics

TL;DR

The first open implementation of the antibody lead optimization framework is introduced, with a single generalist antibody editor trained on 1.66M Observed Antibody Space homolog pairs to propose homolog-like variants of a seed sequence, editing unseen leads zero-shot, without the per-family retraining original approaches require.

Abstract

Antibody lead optimization calls for a small, bounded set of edits to an existing candidate: substitutions, but also insertions and deletions. Edit-based generative models are the only ones that allocate such an edit budget without fixing the edit positions, the edit count, or the output length in advance. However, the existing approaches Edit Flows and EvoFlows did not release code or complete training specifications. Here, we show that both methods follow the same underlying process -- edits firing one at a time, at learned rates, in continuous time -- the pure-jump case of generator matching over finite sequences. With EditJumps we introduce the first open implementation of this framework, with a single generalist antibody editor trained on 1.66M Observed Antibody Space homolog pairs to propose homolog-like variants of a seed sequence, editing unseen leads zero-shot, without the per-family retraining original approaches require. Replicating this system from scratch exposes why open code is essential for generative biology: reconciling published edit distributions required reverse-engineering an undocumented rate-scaling hyperparameter that dictates realized mutation counts. Moreover, we show that published evaluation metrics are highly sensitive to reference sample size, frequently flipping method rankings. We release our full codebase, automated test suite, and configurations at: https://github.com/VisiumCH/editjumps

View source

Similar papers

#machine learning Conference Open access Oct 2026

LadderEdit: Edit-Level Residual Compression for Memory-Efficient Lifelong Editing of LLMs

Lifelong editing of LLMs requires storing thousands of edits after acquisition. A widely used family of approaches attaches one LoRA adapter per edit, which preserves behavior but grows linearly in storage. To address this challenge, we propose LadderEdit, a method that compresses each LoRA adapter after it is acquired...

Xiao-Bing Yu, Pei-Jie Qiu, Jin Yang et al. · 0 citations
Preprint Nov 2026

Tree-Ising: Repairing Language Models with Bit-Level Ising Edits

Weight-space programming, introduced in recent MIT work by Takehi et al., enables a frontier LLM agent to read a network's weights and write a repair. Their GPT-6 Astra agent corrected the reasoning loops of Qwen3.5-0.8B without gradients, using 158 agent turns, 89.7 million tokens, and 50 forward-pass requests. Our me...

Hadi Salloum, Ammar Ali · 1 citation
Preprint Oct 2026

Recursive Self-Improvement on a Budget: Scoring Code Edits Before They Run

A coding agent that rewrites its own scaffold has to choose which proposed edit to expand before that edit has been evaluated, and the search then has to return one agent. Each run fits one linear readout on the parent's steering-split solve rate and eight coordinates of the proposed diff, from 96 scaffold edits labele...

E. Darzi, Sree Bhargavi Balija, Sheng-Feng He · 0 citations
Preprint Nov 2026

IWIBench: Evaluating Gradient-Free Model Driven Intra-weight Injection

Intra-weight injection lets one model change another model's behaviour by editing its weights directly, without gradients. Such edits can range from altering lexical associations to enforcing output formats or updating factual knowledge. Despite growing interest in these approaches, there has been no systematic way to...

Chris von Csefalvay, Pranav Doma, Achyuthan Unni Krishnan et al. · 1 citation
Preprint Nov 2026

Same Fix, Different Weights: Replicating Weight-Space Programming with Claude

Takehi et al. (2026) report that GPT-6 Astra, given the weights of small neural networks as text, can say what they were trained to do, repair them, and write new ones without training. I re-ran the core of that study with Claude Opus 5.5 in ten pre-registered experiments, with untrained-network and permutation control...

Subham Kundu · 0 citations
#machine learning Preprint Nov 2026

What Does Weight-Space Programming Add to Diffusion Editing? Program Construction versus Score-Guided Refinement

Weight-space programming (WSP) lets a language model construct executable weight updates for another network. We study diffusion concept editing through an interface in which programs define training contexts, preservation constraints and update directions, while a shared numerical fitter selects their coefficients. We...

Zitian Gao, Jing-Yi Xu, Zi-Qi Han et al. · 0 citations

Related blog posts

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.