- Rating
- 1(−1)
- Confidence
- 4
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This paper presents a directly trained spiking VLA for robotic manipulation. It combines spiking visual and language encoders, bidirectional competitive fusion, and an action chunk decoder. Experiments on Meta World, LIBERO, and LIBERO Plus report competitive success rates with a compact model and low estimated arithmetic energy.
Critical strengths
-
The compact architecture achieves promising manipulation performance. With approximately 149 million parameters, it reaches 92.4% success on LIBERO and 72.4% on Meta World, supporting the feasibility of directly trained spiking manipulation policies.
-
The fusion mechanism has useful ablation support. Adding competitive routing improves average LIBERO success from 90.2% to 92.4%, with larger gains on Spatial and Long tasks.
Critical weaknesses
-
The distinction from SpikeVLA is not sufficiently established experimentally. SpikeVLA already introduces spiking visual, language, and action components. The present paper emphasizes direct training rather than conversion and targets manipulation, but provides no direct comparison or controlled experiment isolating these differences. Its contribution should therefore be distinguished more precisely from the existing spiking VLA framework.
-
The efficiency evidence does not establish practical deployment gains. Energy is estimated using arithmetic counts and reference operation costs, excluding memory access and non-synaptic overhead. There are no measured latency or hardware energy results, and the encoder ablations change architecture and pretraining alongside the computation type. Consequently, the benefits attributable specifically to spiking computation remain unclear.
Questions for authors
-
What are the specific mechanistic differences from SpikeVLA in spike encoding, temporal integration, multimodal fusion, and action decoding? Beyond replacing conversion with direct training, which changes are essential, and what ablations support them?
-
Could the authors compare against compact ANN policies using INT8 or INT4 quantization under comparable training and evaluation settings, reporting task success, memory, measured latency, and energy? This would clarify whether the proposed spiking mechanisms provide benefits beyond conventional quantization.
10 comments
Weakness 2 and Question 2: Could the authors compare against compact ANN policies using INT8 or INT4 quantization under comparable training and evaluation settings, reporting task success, memory, measured latency, and energy? This would clarify whether the proposed spiking mechanisms provide benefits beyond conventional quantization.
Response: Thank you for the suggestion. We have added comparisons with published quantized VLA policies[2] and architecture-matched experiments to examine whether the benefits of Spike-driven VLA extend beyond low-bit representation.
Table 1. LIBERO success rates (%) and reported memory footprints. Quantized baseline results are taken from QuantVLA.
| Method | Setting | Memory (GB) ↓ | Spatial | Object | Goal | Long | Avg. ↑ |
|---|---|---|---|---|---|---|---|
| π0.5 | FP16 | 4.27 | 98.5 | 99.0 | 97.5 | 93.5 | 97.1 |
| π0.5 + DuQuant | W4A8 | 1.17 | 86.0 | 97.5 | 71.5 | 50.0 | 76.3 |
| GR00T N1.5 | FP16 | 2.02 | 92.0 | 92.0 | 86.0 | 76.0 | 86.5 |
| GR00T N1.5 + DuQuant | W4A8 | 0.74 | 66.0 | 70.0 | 68.0 | 76.0 | 70.0 |
| Spike-driven VLA (Ours) | Spiking | 0.85† | 95.0 | 95.4 | 96.6 | 82.4 | 92.4 |
Spike-driven VLA achieves higher LIBERO success than QuantVLA , with a lower reported memory. To investigate the contribution of spatio-temporal optimization beyond low-bit representation, we further evaluated a single-step variant () and an architecture-matched W8A4 ANN.
Table 2. Ablation results on LIBERO.
| Variant | Avg. success rate (%) ↑ |
|---|---|
| Spike-driven VLA () | 78.6 |
| Matched ANN, W8A4 | 86.2 |
| Spike-driven VLA (full model) | 92.4 |
Reducing the simulation window to decreases success by 13.8%, highlighting the contribution of temporal processing. The full model also exceeds the W8A4 ANN by 6.2%. These findings support the role of joint spatio-temporal optimization in the spiking domain, with performance benefits beyond the tested low-bit ANN representation.
Reference:
[1] Song R, Nie D, Teng S, et al. SpikeVLA: Vision-Language-Action Models with Spiking Neural Networks[J]. arXiv preprint arXiv:2606.27807, 2026.
[2] Zhang J, Hsieh Y, Wan Z, et al. Quantvla: Scale-calibrated post-training quantization for vision-language-action models[J]. arXiv preprint arXiv:2602.20309, 2026.
Which tasks did you evaluate on a physical robot, and what metrics did you use? Could you report the number of trials and the success rate for each task, along with the corresponding baseline results where available?
We thank you for your constructive comments and suggestions, which have been valuable in improving the quality of our work. We provide point-by-point responses below.
Weakness 1 and Question 1: What are the specific mechanistic differences from SpikeVLA in spike encoding, temporal integration, multimodal fusion, and action decoding? Beyond replacing conversion with direct training, which changes are essential, and what ablations support them?
Response: We acknowledge SpikeVLA’s[1] pioneering contribution to integrating spiking neural networks into VLA models. Our method differs from SpikeVLA in three key aspects.
(1) Training paradigm. SpikeVLA combines ANN-to-SNN conversion for perception with PPO training for action control. These stages optimize representation reconstruction and control returns, respectively. However, policy updates do not jointly refine the perception module, limiting task-specific feature adaptation. In contrast, our unified action loss jointly optimizes visual encoding, cross-modal fusion, and action decoding, allowing action errors to directly guide visual feature learning.
(2) Inference latency. Spike-driven VLA directly optimizes action prediction within a fixed (T=4) simulation window, supporting cross-layer spike propagation and asynchronous inference. In SpikeVLA’s conversion-based approach, (T) governs layer-wise ANN activation reconstruction. Under layer-wise blocking execution, reconstruction across (L) layers can require approximately (T\times L) sequential simulation steps. Our end-to-end framework avoids waiting for complete layer-wise activation reconstruction, helping reduce synchronization overhead.
(2) Architecture and scale. SpikeVLA relies on an 8B language backbone for multimodal reasoning. Our method uses MWSF for sparse cross-modal fusion and SpikeACT for parallel action-chunk prediction, with approximately 0.15B parameters in the complete model.
As we have not obtained a complete public implementation or checkpoints of SpikeVLA, we cannot currently provide a controlled comparison on identical tasks. The key differences are summarized below.
*Here, N denotes network depth.