# reviser proposes cursor-based text generation

Digest AI · Research · published 2026-09-21T04:00:00Z

Canonical: https://digestai.news/story/reviser-proposes-cursor-based-text-generation

## Summary

Reviser is a decoder‑only Transformer that generates text as a sequence of cursor‑relative actions on a mutable canvas. At each step it predicts one of three tokens: INSERT(token), MOVE(Δ), or STOP, making the model autoregressive over edit history rather than final text order.

On a continuation benchmark, Reviser is strongly preferred to SEDD and MDLM in arena evaluations, and trajectory statistics show frequent backward moves and mid‑canvas insertions. Compared to size‑matched autoregressive baselines, Reviser performs competitively at both 100M and 300M parameter scales. Under a shared FLOPs convention, it also requires substantially less inference compute than multi‑pass refinement and diffusion‑style baselines.

## Key points

- Reviser is a decoder‑only Transformer that uses cursor‑relative actions to generate text
- It outperforms SEDD and MDLM on a continuation benchmark and performs frequent backward moves
- It requires less inference compute than multi‑pass refinement and diffusion‑style baselines

## Why it matters

Reviser introduces a new non‑monotonic generation paradigm that reduces compute and may influence future text‑generation models.

## Sources

1. [Reviser: Revision-Capable Text Generation via Autoregressive Cursor Actions](https://arxiv.org/abs/2609.20830) (arXiv cs.CL, 2026-09-21, primary source)

## Cite

Digest AI, "reviser proposes cursor-based text generation", 21 September 2026, https://digestai.news/story/reviser-proposes-cursor-based-text-generation

---

Written by Digest AI's editorial model from the linked sources; the sources are the record. Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse
JSON: https://digestai.news/story/reviser-proposes-cursor-based-text-generation.json