- Rating
- 2
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 1
Read review
Paper summary
The paper extends text diffusion with token insertion and deletion, so output length can change during generation. It also trains ASR correction models on errors produced by a TTS-to-CTC pipeline and introduces faster decoding and rescoring methods. Experiments show better code infilling than the tested diffusion baselines and a useful ASR accuracy-speed trade-off on LibriSpeech. General text generation is comparable only with additional length initialization.
Critical strengths
Insertion, deletion, and substitution address a real limitation of fixed-length diffusion, with clear gains on the reported code-infilling tasks.
Critical weaknesses
-
Appendix B.1 uses weighted cross-entropy without noise-level conditioning, and evaluates neither iterative denoising nor the count head. In this setup, the model acts as a parallel edit predictor. A matched non-diffusion editor is not compared, so faster decoding than an autoregressive corrector does not isolate what the diffusion formulation adds. This weakens the connection between the paper’s two main contributions.
-
In Table 6, outputs average only 6.0 tokens without forced insertions; two initial updates raise this to 14.7, and three to 29.0. MAUVE rises from 0.46 to 0.84 after two updates. Appendix A.3 also reports count underestimation. This leaves the model’s ability to learn the output-length distribution unconvincing, which matters because flexible length is the central contribution.
Questions for authors
In the reported ASR system, which training or inference operation distinguishes Flex ASR-DiffLM from a non-diffusion parallel edit model trained on the same TTS-CTC pairs?
How were the two forced initialization updates selected, and what existing evidence shows that this choice transfers across target-length ranges without retuning?
20 comments
But one-step correction is not specific to Flex ASR-DiffLM. A DiffLM includes iterative denoising, a noise schedule during training, noise conditioning and the clean-token predictor . For speech recognition, we use all our diffusion LMs (USDM, Flex-DiffLM, ASR-DiffLM and Flex ASR-DiffLM) as single-step correctors, which only need . We do not use the count head for any of them.
For iterative denoising, our results show that it does not give an improvement on LibriSpeech.
Table B. USDM (25 epochs) iterative joint decoding on LibriSpeech dev-other. WER [%] for different numbers of denoising iterations and initial noise levels.
| Iterations | ||||
|---|---|---|---|---|
| 1 | 4.60 | 4.63 | 4.63 | 4.65 |
| 2 | 4.59 | 4.60 | 4.62 | 4.62 |
| 3 | 4.59 | 4.60 | 4.60 | 4.62 |
| 4 | 4.58 | 4.59 | 4.60 | 4.60 |
| 5 | 4.59 | 4.60 | 4.60 | 4.60 |
This means that for ASR, we are only interested in the difference in how is trained.
From the training perspective, standard discrete DiffLMs are also close to parallel denoising models. MDLM training is masked language modeling like BERT with a sampled masking rate and USDM predicts the clean tokens from a sequence with uniformly replaced tokens in the same way. Our ASR-specific DiffLMs keep this setup and only replace the synthetic corruption with TTS-CTC errors.
Thank you very much for the review and for the two very important questions. Below, we will provide additional experiments and we believe that the new results should address the concerns you raised.
Yes, we agree. In our ASR-specific DiffLM, there is no substantial difference that distinguishes Flex ASR-DiffLM from a non-diffusion parallel edit model trained on the same TTS-CTC pairs with our augmented canvas. It was intended as a first step to introduce real ASR errors into the forward process. Our framing in the submission was not clear enough here and we have revised Section 1, Section 4, Section 5 and Appendix C to make it clearer.
Our motivation was a controlled comparison. All four models use the same architecture and the same decoding methods and they differ only in the sequence length and in the corruption:
Table A (Table 7). Relation between our DiffLM formulations for speech recognition.
Our solution to the fixed-length problem is the augmented canvas, while training on real ASR errors is our solution to synthetic corruption. Combining the augmented canvas with real ASR errors shows the largest improvement over standard USDM (4.22 to 3.88 with the best decoding method of each model, CTC trained without TTS data, Table 5).