- Rating
- 2
- Confidence
- 2
- Informativeness
- 3
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
The paper propose MORPHEUS-2.0, an encoder-only time-series foundation model that use single variable-time grid to represent targets and covariates. The MORPHEUS-2.0 achieve 0.689 relative MASE with 81M in GIFT-EVAL and its multiscale counterpart attains the lowest relative MASE compared to other system on benchmark.
Critical strengths
- MORPHEUS-MULTISCALE achieves the lowest relative MASE on GIFT-Evaluation, while its 81M-parameter variant also delivers competitive performance.
- The paper is well written, with clear explanations and intuitive illustrations that make the proposed method easy to understand.
- The prediction-length randomization and four-stage synthetic-to-real curriculum are thoughtfully designed, demonstrating strong consideration of practical forecasting challenges.
Critical weaknesses
-
Although the paper emphasizes the computational benefits of dual-axis attention and reduced activation memory, it does not systematically evaluate inference efficiency. MORPHEUS-MULTISCALE’s three encoders with 247M parameters may incur substantial overhead; reporting latency, peak GPU memory, throughput, and FLOPs would better characterize the efficiency–accuracy trade-off.
-
The performance improvements of MORPHEUS-MULTISCALE may stem from ensembling rather than multiscale modeling. As shown in Appendix D4, combining three encoders with the same patch width already reduces relative MASE from 0.689 to 0.669, while using different patch widths further improves it to 0.653. This suggests that much of the improvement comes from standard ensembling, with the distinct contribution of multiscale fusion being relatively modest.
-
Table 2 demonstrates the benefits of attention factorization and learnable gating, but it does not establish whether simultaneous processing of both axes outperforms the conventional alternating axial attention discussed in Section 4.3. A comparison with an alternating axial attention baseline under matched parameter counts, model depth, and training budgets would therefore strengthen the paper.
Questions for authors
-
Could the authors provide additional zero-shot evaluation results on independent benchmarks beyond GIFT-Eval?
-
Could mixing-weight visualizations clarify whether the context-conditioned weights align meaningfully with signal characteristics, forecast horizons, and sampling frequencies?
7 comments
W1 - Inference efficiency
Appendix E7 now reports this. Both models run on the same A100 at fp32 on identical GIFT-Eval requests, timing every encoder, the mixing head and pooling. Every timed forecast matched the evaluated output, so cost and accuracy refer to the same predictions.
On short-horizon configurations, Morpheus-Multiscale needs about 3.5× the FLOPs of Morpheus-2.0, 4.5× the latency and 2.2× the peak memory, with about 4× lower throughput. This overhead is the price of accuracy: it buys 7.8% lower MASE and 12.2% lower CRPS, with both lower on almost every configuration. The cost comes from running several encoders, the narrow-patch ones on longer sequences, and is fixed by request shape rather than data (FLOPs 3.4–3.7×, memory 2.2–2.3×), so it is known in advance. In high-stakes forecasting, a forecast error costs far more than the extra seconds of GPU time needed to avoid it, and that is the trade the multiscale model makes. Dual-axis attention keeps each encoder compact, with query–key pairs instead of and under half the activation memory (Section 7.1).
We therefore release two operating points. Morpheus-2.0 suits high-volume forecasting and is more accurate than the larger Chronos-2. Morpheus-Multiscale targets decisions where errors cost more than compute, such as energy dispatch, inventory and capacity planning, which usually run at planning intervals rather than under latency limits.
| Morpheus-Multiscale relative to Morpheus-2.0 (Appendix E7) | Change |
|---|---|
| FLOPs per request | 3.5× |
| Latency, one request per call | 4.5× |
| Throughput, four requests per call | 4× lower |
| Peak GPU memory, four requests per call | 2.2× |
| Relative MASE, short horizons | −7.8% |
| Relative CRPS, short horizons | −12.2% |
W2 - Ensembling versus multiscale modeling
Appendix D4 addresses this question, and we have expanded it with equal-weight and best-per-horizon controls (Table 13, below). They show that ordinary ensembling saturates, and that the remaining gain needs both multiple resolutions and context-dependent weighting.
Ensembling at one resolution reaches its limit quickly. Three patch-4 encoders improve MASE from 0.689 to 0.669, and that is all ensembling at one resolution can contribute; spending the same budget on one larger patch-4 encoder is far worse. Adding resolutions goes further. Replacing two of the patch-4 copies with patch-2 and patch-8 encoders improves the combination on both metrics, even though each new encoder is individually weaker than the copy it replaces, because each width sees structure the others cannot resolve: in Table 3, patch 2 wins on oscillations and wide schemas, patch 4 on impulses and noisy series, and patch 8 on persistent levels. No single width, and no ensemble of one width, covers all of these, which is why multiple resolutions are required.
Equal averaging cannot fully exploit this, because where one width is right the others pull the forecast away from it. The mixing head reads the context and weights the right width, lowering MASE to 0.653, whereas choosing the best encoder per horizon, even with hindsight, does worse than equal averaging.
Together, resolution diversity and context weighting account for 2.4% of the 5.2% gain over the standalone model, nearly half. At the top of GIFT-Eval that is substantial: it exceeds the spread across the five systems from Falcon-2.0 to Toto-2.0-2.5B-FT in Table 1.
| Combination of encoders (Table 13) | Params | MASE | CRPS |
|---|---|---|---|
| Patch 2, single encoder | 84.3M | 0.696 | 0.507 |
| Patch 4, single encoder | 81.4M | 0.689 | 0.522 |
| Patch 8, single encoder | 80.1M | 0.708 | 0.526 |
| Patch 4, single encoder scaled to multiscale budget | 247M | 0.838 | 0.602 |
| Three patch-4 encoders, equal weights | 245M | 0.669 | 0.492 |
| Widths 2, 4, 8, best encoder per horizon | 246M | 0.667 | 0.494 |
| Widths 2, 4, 8, equal weights | 246M | 0.661 | 0.490 |
| Widths 2, 4, 8, mixing head | 247M | 0.653 | 0.485 |
Summary of revisions
Thank you for the review. We added everything you requested: measured inference cost, an alternating axial baseline at matched budget, equal-weight and per-horizon controls that separate ensembling from the multiscale gain, and an analysis of the learned mixing weights on GIFT-Eval. Results on an additional benchmark will follow in this thread. With the revisions, Morpheus-Multiscale reaches 0.642 relative MASE and 0.454 relative CRPS, the lowest of any published system on both (Table 1).