Pixel-space generative Transformers typically use one token per image patch, so capturing finer detail requires more tokens and makes global attention increasingly expensive. We observe that finer detail does not necessarily require more global communication: low-frequency structure benefits from global exchange, while high-frequency detail can be modeled locally. Based on this insight, we introduce SpecSlot, a pixel-space rectified-flow Transformer that decouples output detail from the length of the global attention sequence. SpecSlot expands each spatial anchor into multiple slots, each assigned a disjoint band of DCT coefficients; this separates anchor density from spectral subdivision. Only the lowest-frequency slot from each anchor participates in global attention, while all slots interact locally. This allows the model to increase spectral detail without enlarging the global sequence. A spatial head further adds a local correction to the spectral prediction, while the generative state, target, and loss remain in pixel space. On ImageNet-256, SpecSlot achieves a comparable final FID to a FLOPs-matched JiT control while converging faster. On ImageNet-512, it improves FID by 2.84 over the control with 3.9% fewer FLOPs, and comes within only 0.69 FID of JiT-B/16 at 38.2% lower cost. On text-to-image generation, SpecSlot surpasses JiT-B/8 on DPG-Bench by 1.06 at 40.6% lower forward cost.
Figures & tables
Figure 2: SpecSlot data flow. Each anchor expands into K slots. G blocks process lowest-radial slot-0 carriers across anchors with context states; L blocks mix all slots within each anchor. The final slot states decode a lapped-DCT spectral base and an anchor-aligned spatial correction, summed in pixel space.
Figure 3: SpecSlot allocation. A patch grid spends all N tokens in one global sequence; splitting N=WK into W spatial anchors and K spectral slots keeps global attention at W carriers.
Figure 4: Mean FID-10K under the aligned protocol.
Model
Tokens
B
CFG
FID ↓
IS ↑
GFLOPs ↓
img/s ↑
ImageNet-256
JiT-B/16
256
–
2.90
7.51
174.81
50.45
33.08
JiT-B/8
1,024
–
2.90
4.34
225.72
218.50
8.15
SF-B/16
256
–
2.90
7.81
172.35
50.85
32.89
JiT-CM
1,024
–
2.90
6.68
182.56
124.24
8.94
SpecSlot (W256/K2)
512
512
5.00
7.27
219.49
82.63
21.32
Table 1: ImageNet quality and compute. SpecSlot rows report the best swept B and CFG per allocation (SF-B/32 at ImageNet-512: CFG 9.50). GFLOPs are conditional forward costs; img/s is peak Heun-50 throughput; – indicates that SpecSlot’s B parameter does not apply. SpecSlot gives two nondominated operating points between JiT’s low- and high-cost settings, and its margin over the compute-matched control is larger at ImageNet-512 than at ImageNet-256 in these comparisons. JiT models are retrained for 120 epochs under the aligned recipe.
Figure 5: Paired ImageNet-256 generations with the same class label and noise draw: SpecSlot W256/K4 and JiT-CM.
Model
CFG
GFLOPs ↓
DPG ↑
GenEval ↑
JiT-B/16
2.9
50.71
59.22
24.89
JiT-B/8
2.9
218.77
61.15
29.78
SpecSlot (W256/K4)
4.0
129.86
62.21
26.99
Table 2: Text-to-image generation at 256×256 after training 400 epochs using the same prompt sets and 50-step Heun sampler. DPG-Bench [ 28 ] and GenEval [ 20 ] report overall scores in percent; forward GFLOPs include the trainable text adapter.
Route
Cross-anchor image states
FID ↓
IS ↑
GFLOPs ↓
img/s ↑
GGLL
one carrier per anchor
6.59
215.20
129.20
9.06
GLGL
one carrier per anchor
6.95
218.95
129.20
9.00
Dense-global-CM
all image slots
6.81
210.20
128.96
5.80
Full-global
all image slots
5.84
237.54
225.82
5.11
Table 3: Communication schedules at fixed W256/K4/B1536 ; efficiency columns come from one profiler run.
Figure 6: Evidence timing and routed-communication probes. Panels (a)–(c) report the data-to-noise ratio crossing time per radial band for paired W256/K8, W256/K4, and W1024/K2 checkpoints, frequency-group recovery, and route sensitivity. The route interventions remove cross-anchor carrier edges or local cross-band edges inside the shared Transformer while retaining carrier self edges and context reads.
K
Velocity MSE
Effective slots/ K
Effective local keys
Detail → carrier
2
.10670
.803
.729
.330
4
.10549
.620
.682
.312
8
.10497
.495
.621
.275
Table 5: Effective use of additional spectral slots at fixed W=256 and evidence width B=1536 . Values are means over paired teacher states. Effective slots/ K is the normalized entropy effective rank; definitions are in Appendix E.2 .
Figure 7: Output-branch probes. Panels (a)–(c) show normalized branch self-energies and their signed cross term, spatial self-energy share by radial band, and relative velocity MSE under spatial or spectral scaling over t∈{0.05,…,0.90} . The energy panel shows 95% bootstrap intervals; the other panels show means.
Global carrier
FID ↓
IS ↑
Slot 0 (lowest radial)
6.59
215.20
Slot 1
6.91
220.38
Slot 2
7.21
203.90
Slot 3
6.97
214.87
Table 7: Carrier-identity ablation at fixed W256/K4/B1536 : only the slot with cross-anchor access changes.
Appendix figures & tables9 assets
Supplementary material from the paper’s appendix.
Appendix
Quantity
Value
Image side H
256
Input cell / stride s
16 / 16
Spatial anchors W
256
Slots per anchor K
4
Image slots N=WK
1,024
Hidden / FFN width / heads / blocks
768 / 3,072 / 12 / 12
Appendix
Table 8: Implementation and training contract of the class-conditional W256/K4 models. Resource rows are measured on one H800 with BF16 whole-network TorchInductor at profiler batch 128; the ImageNet-512 GFLOPs and throughput are the main-table measurements, and its peak-memory row is the logged training peak per GPU at batch 128.
Quantity
Value
Backbone configuration
W256/K4/B1536 at 256×256 (Table 8 )
Text encoder
frozen OpenCLIP ViT-L/14, 77-token states
Text adapter
32-query single-layer cross-attention
Training data
BLIP3o long/short-caption + JourneyDB shards
Training epochs / GPUs
400 / 8 (64 images per GPU)
Optimizer / learning rate
AdamW / 2×10−4
Appendix
Table 9: Text-to-image implementation and training contract (Section 4.3 ).
Route
Heun
Evals/image
GFLOPs/eval
TFLOPs/image
ms/image
img/s
GGLL
20
73
128.59
9.39
47.56
21.02
Dense-global-CM
20
73
128.96
9.41
69.89
14.31
Full-global
20
73
225.22
16.44
80.99
12.35
GGLL
30
113
128.59
14.53
73.09
13.68
Dense-global-CM
30
113
128.96
14.57
109.20
9.16
Full-global
30
113
225.22
25.45
124.83
8.01
Appendix
Table 11: Fixed-batch sampling latency on one H800. “Evals/image” is the actual network-call count under classifier-free guidance; “TFLOPs/image” is the analytic cumulative model-forward compute.
Figure 8: Sampling latency as the evaluation budget increases. The x-axis gives the profiler-observed network evaluations per image; the nominal Heun intervals appear in parentheses.
Allocation
Training seed
FID-50K ↓
IS ↑
W256/K2
12345
7.2191
214.2811
12346
7.0054
237.4801
12347
7.6025
206.7721
12348
7.3332
226.7448
12349
7.1983
212.1924
W256/K4
12345
6.4266
232.6772
Appendix
Table 12: Per-training-seed ImageNet-256 evaluation on 50K generated images per checkpoint. FID and IS are reported for training seeds 12345–12349.
Allocation
t⋆ for successive radial groups
W256/K8
{0.382,0.836,0.880,0.904,0.922,0.933,0.943,0.959}
W256/K4
{0.468,0.891,0.928,0.951}
W1024/K2
{0.549,0.936}
Appendix
Table 14: Empirical DNR frontiers from the lapped-frame probe. Entries list the crossing times t⋆ in increasing radial-group order.
Figure 9: Selected paired generations on ImageNet-256. Within each pair, JiT-CM is shown on the left and SpecSlot W256/K4 on the right; both use the same class and noise draw.
Figure 10: Additional text-to-image samples generated by SpecSlot W256/K4 from DPG-Bench prompts (first part).
Figure 11: Additional text-to-image samples generated by SpecSlot W256/K4 from DPG-Bench prompts (continued).