- Rating
- 2
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This paper investigates long-session degradation in full-duplex speech models, including declining responsiveness, conversational quality, and speech quality. The authors propose privileged-context self-distillation to improve behavioral stability and a Two-Resolution Context (TRC) architecture that preserves recent context while compressing older dialogue. Experiments on three model architectures show that the proposed approach generally improves stability within the 10-minute training horizon and, when combined with TRC, better preserves performance in 20-minute sessions. TRC also reduces KV cache growth by 91.7%–95.8%. However, the gains vary across models, responsiveness issues remain in some cases, and compressed memory can impair information recall. Overall, the paper provides a useful evaluation framework and a promising approach to improving long-session stability.
Critical strengths
-
The paper evaluates degradation in responsiveness, conversational quality, and speech quality across multiple full-duplex speech models. This provides a useful evaluation framework and highlights an important but underexplored problem in sustained speech interaction.
-
The proposed privileged-context self-distillation and Two-Resolution Context (TRC) methods improve long-session stability in several settings. Experiments show that TRC reduces KV cache growth by 91.7%–95.8% and better preserves performance in 20-minute sessions, supporting the practical value of the approach.
Critical weaknesses
-
Insufficient validation under realistic full-duplex interactions. The user simulator does not initiate interruptions or backchannels, leaving the proposed methods untested under important real-world interaction scenarios. This weakens the external validity of the evaluation and makes it difficult to conclude that the methods reliably address long-session degradation in practical settings. This is a significant concern for acceptance because realistic full-duplex behavior is central to the paper's motivation.
-
Incomplete and model-dependent mitigation of degradation. TRC does not consistently improve responsiveness across architectures, and context compression can impair long-range memory recall. These results suggest that the proposed methods do not comprehensively resolve long-session degradation. Without stronger evidence of robustness across models and interaction conditions, the paper's claims of general effectiveness are not fully supported.
Questions for authors
-
How do the authors justify the effectiveness of their methods when the simulator does not include user-initiated interruptions or backchannels?
-
Why does TRC fail to consistently improve responsiveness, particularly for MiniCPM-o 4.5? What evidence supports its effectiveness across different architectures?
-
How much information is lost through TRC compression, and how does this loss affect long-session conversational quality?
-
How can the authors demonstrate that the proposed methods generalize beyond the 10-minute training horizon, given the limited evaluation up to 20 minutes?
-
Have the authors validated the LLM-based conversation quality scores and ASR-based WER against human judgments to confirm that the reported improvements reflect actual user-perceived quality?
6 comments
W1/Q1: Validation under realistic full-duplex interaction (2/3)
Mitigation of the degradation. We next evaluate Ours (TRC) in 20-minute sessions with GPT-Live-1. As an additional setting, we also evaluate Ours (TRC, 30-s teacher) on MiniCPM-o 4.5, a variant of TRC trained on the same trajectories, with the teacher's context window for computing the targets shortened from up to 150 s to 30 s. Each cell reports the 8–10 / 18–20 min intervals; – marks intervals beyond the 10-minute Base sessions or below the reporting thresholds.
| Model | Method | RT (%) ↓ | TS (%) ↓ | FTO (s) | CQ ↑ | WER (%) ↓ | Spk sim. ↑ |
|---|---|---|---|---|---|---|---|
| PersonaPlex | Base | 35.8 / – | 36.7 / – | 0.45 / – | 1.16 / – | 53.1 / – | 0.224 / – |
| Ours (TRC) | 12.6 / 8.7 | 3.3 / 10.0 | 0.23 / 0.23 | 3.91 / 3.28 | 11.7 / 18.2 | 0.599 / 0.537 | |
| MiniCPM-o 4.5 | Base | 55.3 / – | 76.7 / – | – / – | 4.75 / – | 8.9 / – | 0.735 / – |
| Ours (TRC) | 66.7 / – | 96.7 / 100.0 | – / – | – / – | – / – | – / – | |
| Ours (TRC, 30-s teacher) | 7.8 / 9.1 | 0.0 / 3.3 | 0.96 / 0.84 | 4.65 / 4.60 | 4.6 / 4.7 | 0.772 / 0.779 |
On PersonaPlex, Ours (TRC) mitigates the degradation through 20 minutes with GPT-Live-1 as well. It keeps content and speech quality well above the late-session Base level and terminates far fewer sessions: 10.0% by 20 min, against 36.7% of Base sessions by 10 min. Its response timeouts are more frequent than with our simulator, but they do not increase further beyond the training horizon.
On MiniCPM-o 4.5, in contrast, Ours (TRC) does not mitigate the degradation: 96.7% of its sessions terminate by 10 min, more than for the Base model. This reveals that, on this model, the configuration used in the paper does not generalize well to interaction with more active full-duplex behaviors, a limitation already partly visible with our simulator. Changing the configuration, however, substantially improves the mitigation. With Ours (TRC, 30-s teacher), response timeouts stay below 10%, only one session (3.3%) terminates through 20 minutes, and content and speech quality remain stable.
Overall, the degradation of the base models reappears when the user is a real-time full-duplex system that backchannels and interrupts, which suggests that our findings on degradation are not specific to the calm turn-taking or the logical-time protocol of our simulator. As for mitigation, these results suggest that TRC itself can generalize to such interaction, but that applying it with the same configuration across models may not be sufficient. They also point to both a need and room for adapting the configuration to each model, with attention to the trade-offs this could introduce.
W1/Q1: Validation under realistic full-duplex interaction (3/3)
(2) Full-duplex behaviors of the adapted models
We evaluated the base models and Ours (TRC) on Full-Duplex-Bench v1.0 (Lin et al., "Full-Duplex-Bench: A Benchmark to Evaluate Full-duplex Spoken Dialogue Models on Turn-taking Capabilities," ASRU 2025). We used its public implementation for transcription and metric computation.
- Prompts: PersonaPlex uses the task-specific prompts specified in the benchmark repository. The benchmark specifies no configuration for MiniCPM-o 4.5 or Raon-SpeechChat, so we applied the same prompts to them.
- Decoding: All models use the default sampling settings of their released inference code. The base and TRC variants of each model share identical settings.
- Judge: The user-interruption judge is GPT-4o (gpt-4o-2024-08-06), because GPT-4-turbo, used in the original script, has been retired.
- Raon-SpeechChat: Its technical report (Kim et al., "Raon-Speech Technical Report," 2026, Appendix G) evaluates this benchmark with an internal evaluator that uses different turn-detection criteria. Our absolute values for this model therefore differ substantially from the reported ones. The comparison below uses the same public protocol for both variants.
| Model | Method | Pause Synth. TOR ↓ | Pause Candor TOR ↓ | BC TOR ↓ | BC Freq ↑ | BC JSD ↓ | Turn TOR ↑ | Turn Lat. (s) ↓ | Intr. TOR ↑ | Intr. GPT ↑ | Intr. Lat. (s) ↓ |
|---|---|---|---|---|---|---|---|---|---|---|---|
| PersonaPlex | Base | 0.394 | 0.454 | 0.255 | 0.122 | 0.735 | 0.958 | 0.29 | 0.915 | 4.60 | 0.19 |
| Ours (TRC) | 0.255 | 0.329 | 0.273 | 0.078 | 0.781 | 0.899 | 0.31 | 0.925 | 4.48 | 0.41 | |
| MiniCPM-o 4.5 | Base | 0.212 | 0.588 | 0.218 | 0.017 | 0.910 | 0.958 | 0.36 | 0.870 | 2.84 | 0.50 |
| Ours (TRC) | 0.058 | 0.120 | 0.073 | 0.011 | 0.945 | 0.849 | 0.65 | 0.710 | 2.15 | 0.46 | |
| Raon-SpeechChat | Base | 0.854 | 0.917 | 0.473 | 0.198 | 0.660 | 1.000 | 0.00 | 1.000 | 2.92 | 0.29 |
| Ours (TRC) | 0.745 | 0.875 | 0.309 | 0.143 | 0.711 | 1.000 | 0.01 | 1.000 | 2.94 | 0.26 |
The adapted models still take the turn in most cases when the user yields it and respond to most user interruptions (TOR of at least 0.849 and 0.710, respectively), but the adaptation shifts these behaviors to some extent. Two changes are shared by all three models. They handle user pauses better, with lower TOR on both pause subsets (e.g., 0.588 → 0.120 on Candor for MiniCPM-o 4.5), and produce backchannels less frequently and with timing further from that of human listeners (higher JSD). Beyond these shared changes, PersonaPlex and MiniCPM-o 4.5 take the turn less reliably when the user yields it (smooth turn-taking TOR 0.958 → 0.899 and 0.958 → 0.849), and MiniCPM-o 4.5 shows a marked decline in handling user interruptions (TOR 0.870 → 0.710; GPT score 2.84 → 2.15).
The decline in interruption handling can be expected from the trajectory design. The simulated user never interrupts the model (Appendix B.5), so interruption handling receives no supervision during training. In contrast, the models' own backchannels and floor-taking were left unconstrained during trajectory construction, yet these behaviors shifted as well. This suggests that additional training on a small dataset collected with a limited type of user simulator is cost-efficient but can measurably shift a model's full-duplex behavior beyond what the training targets, in both favorable and unfavorable ways, so such adaptation warrants caution. Examining how this effect changes when trajectories are generated with a full-duplex user simulator such as GPT-Live-1, or when training draws on external full-duplex dialogue data, is an important direction that we leave for future work. We will include this discussion in the limitations.
In summary, these two evaluations suggest that our methods are not confined to the calm turn-taking of our main protocol. With a user that initiates backchannels and interruptions, TRC can still mitigate long-session degradation, and none of the full-duplex behaviors evaluated on the benchmark breaks down after the adaptation. Both evaluations, however, also show that the outcome depends on how the adaptation is configured: the training configuration needs to be adapted to each model, and the user behaviors covered by the training trajectories affect full-duplex behaviors that the training does not directly target.
We appreciate the careful review and constructive feedback. Based on the points raised in the review, we conducted additional experiments to strengthen our evaluation. We address each point below. All additional results, evaluation settings, and discussions in this response will be added to the Appendix of the revised paper.
W1/Q1: Validation under realistic full-duplex interaction (1/3)
We agree that the main protocol leaves the methods untested under user-initiated interruptions and backchannels. We address this with two additional evaluations. First, we repeated the long-session evaluation with a full-duplex user that initiates these behaviors, to test whether the mitigation holds under realistic interaction. Second, we evaluated the adapted models on a full-duplex benchmark, to test whether they still handle behaviors such as user interruptions.
(1) Long-session evaluation with a full-duplex user
We use OpenAI's GPT-Live-1, a commercial full-duplex API operating in real time, as the user. Unlike our simulator, GPT-Live-1 determines its own response timing and initiates backchannels and interruptions. The setting is as follows.
Degradation of the base models. We first compare the Base models under our simulator and GPT-Live-1. Each cell reports the 0–2 → 8–10 min intervals, and the values for our simulator are from Table 1. As in the paper, we apply the reporting thresholds of Appendix A.4 to avoid unstable estimates from too few observations; – marks intervals that do not meet them.
With GPT-Live-1, the degradation of both models largely follows the pattern observed with our simulator, while its severity and onset differ. PersonaPlex loses responsiveness, content quality, and speech quality more severely than with our simulator. MiniCPM-o 4.5 again degrades mainly in responsiveness while keeping its content quality, but the decline starts earlier: 36.9% of response opportunities already time out at 0–2 min, and 76.7% of sessions terminate by 10 min. The degradation found in our evaluation is therefore not confined to calm turn-taking; it also arises when the user backchannels and interrupts in real time.