mlsanityData to 14 Oct 2026

TimpaTeks: Automatic In-place Text Sequence Modification via Diffusion Language Model Steering

arXiv 2606.08408 ↗First posted 7 Jun 2026cs.CL cs.AIcited by 0

On the map: Diffusion models → Discrete and text diffusion

also in 1 more
Abstract (arXiv)

We extend activation steering to diffusion language models (DLMs) and study a novel problem that arose due to the inference mechanism of DLMs: Modifying a text in-place to manifest a different concept. We propose TimpaTeks, an automatic in-place text modification mechanism using DLMs. Experiments on IMDB movie reviews (sentiment) and a synthetic Cats and Dogs Dataset (arbitrary, more unconventional concept steering) show that TimpaTeks provides a feasible novel mechanism to steer diffusion language model outputs in-place. TimpaTeks enables in-place modification while simultaneously lowers sentence perplexity and retaining the original sentence structre without the need of instruction tuned models. TimpaTeks is also computationally cheaper than prompt-based DLM steering, as it performs denoising in-place rather than constructing an additional prompt-conditioned output sequence.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Discrete and text diffusion, one of 14 directions in Diffusion models. Builds on 7 of its 12 references that are in the map. First posted on arXiv 7 Jun 2026.

Builds on

7 of 12 references.

Closest ancestors · 2

Same direction, newest first

  1. 2025Activation Steering for Masked Diffusion Language Models30 Dec
  2. 2024Controlling Language and Diffusion Models by Transporting Activations30 Oct

Other references · 5

  1. 2023Steering Llama 2 via Contrastive Activation Additioncited by 1,085
  2. 2023Steering Language Models With Activation Engineeringcited by 922
  3. 2024Programming Refusal with Conditional Activation Steeringcited by 206
  4. 2024Analyzing the Generalization and Reliability of Steering Vectorscited by 143
  5. 2025LinEAS: End-to-end Learning of Activation Steering with a Distributional Losscited by 14

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026ILRR: Inference-Time Steering Method for Masked Diffusion Language Models
  2. 2026Steering Without Breaking: Mechanistically Informed Interventions for Discrete Diffusion Language Models
  3. 2026DLM-SWAI: Steering Diffusion Language Models Before They Unmask
  4. 2026Diffusion Language Models: An Experimental Analysis
  5. 2024Scaling Diffusion Language Models via Adaptation from Autoregressive Models
  6. 2026In-Place Instruction Following in Diffusion Language Models
  7. 2026Remask, Don't Replace: Token-to-Mask Refinement in Diffusion Language Models
  8. 2025A Survey on Diffusion Language Models