Unified Multimodal Models (UMMs) aim to support visual understanding and generation within a single model. Existing post-training methods either use understanding as a fixed teacher to improve generation, or rely on ground-truth trajectories to supervise both understanding and generation. Prism instead derives bidirectional supervision from self-generated interleaved rollouts, constructing training targets according to the reliability of their text and image chains. We propose Prism, a bidirectional self-teaching post-training method for UMMs. Like a prism decomposing mixed light, Prism decouples a coupled interleaved trajectory into a text chain and an image chain, assesses the reliability of each chain, and constructs a corresponding training target from their reliability states. Experiments on both SenseNova-U1 and BAGEL show that Prism improves both understanding and generation. On the generation side, GenEval improves from 85.8 to 89.2 (+3.4), and DPG improves from 86.1 to 87.3 (+1.2) on SenseNova-U1. On the understanding side, the total MME score improves from 2282.6 to 2311.6, and MMVP improves from 58.0 to 62.0. Consistent improvements are observed on BAGEL. These results highlight that Prism enables coordinated improvements in generation and understanding by constructing reliability-guided self-training targets from self-generated trajectories, without relying on an external verifier or externally annotated corrected trajectories during target construction.
Figures & tables
Figure 1: Capabilities of Prism. Prism improves both understanding and generation through self-teaching on interleaved text-image trajectories. The base model shows redundant steps on step 2 in the cooking tutorial, while Prism produces a coherent and correct tutorial. It also improves performance on standalone understanding and generation tasks.
Figure 2: Difference between Prism and previous post-training methods. (a) Some prior methods are data-driven or rely on external supervision or verifiers. (b) Others improve generation through understanding in a fixed direction. (c) Prism instead enables bidirectional improvement between understanding and generation without external supervision or verifiers.
Figure 3: Prism’s position by ground-truth supervision and improved capabilities.
Figure 4: Overview of the Prism. Given a text prompt, Prism first uses the UMM to roll out an interleaved trajectory. It then uses the UMM in a judging role to assess the reliability of the text chain and image chain. Based on the resulting status, Prism selects the corresponding processing action and loss function. All UMM instances are initialized from the same base model. Only the rollout UMM is trainable, while the UMMs used for judging and action selection remain frozen.
Table 2: Prism improves multimodal understanding performance. We evaluate the effectiveness of Prism across understanding benchmarks covering general knowledge, subject knowledge, and others.
Method
FDT
w/o Tie
w/ Tie (0)
w/ Tie (.5)
SenseNova-U1 + Prism
74.93%
76.65%
66.79%
73.22%
SenseNova-U1
72.03%
72.91%
62.55%
69.66%
GPT-4o+DALL-E3
63.79%
66.92%
56.36%
64.25%
Human
61.10%
64.66%
51.90%
61.76%
Gemini1.5+Flux
38.35%
39.04%
31.23%
41.24%
BAGEL + Prism
34.70%
31.66%
25.80%
35.05%
Table 3: OpenING pairwise evaluation. Left: Win rates under different tie-handling settings. Right: Zero-image rates, i.e., outputs with no generated image. Lower is better.
Figure 5: Improved generation. We use SenseNova-U1 as the baseline model. The model trained with Prism shows more accurate adherence to prompt, including spatial relations, object counts, and style. The underlined text highlights the generation differences under complex instructions.
Model
MME
MMBench
MMMU
MMStar
MMVP
GenEval
DPG
BAGEL
2361
84.6
53.7
65.0
69.3
76.5
84.0
+ SRUM
2350 (-11)
84.4 (-0.2)
53.4 (-0.3)
64.7 (-0.3)
70.0 (+0.7)
77.4 (+0.9)
85.2 (+1.2)
+ UniCorn
2337 (-24)
84.1 (-0.5)
54.5 (+0.8)
65.0 (+0.0)
70.0 (+0.7)
80.3 (+3.8)
86.2 (+2.2)
+ Prism
2371 (+10)
88.4 (+3.8)
55.0 (+1.3)
69.6 (+4.6)
70.0 (+0.7)
78.3 (+1.8)
85.0 (+1.0)
Table 4: Compared with similar methods on BAGEL, our method can simultaneously improve both understanding and generation, demonstrating bidirectional enhancement capability.
Figure 6: Improved understanding. We use SenseNova-U1 as the baseline model. The model trained with Prism demonstrates more accurate responses on diverse multimodal multiple-choice tasks, including code reasoning, visual counting, numerical reasoning, and commonsense understanding. For more examples, please refer to the Appendix G
Loss Terms
Understanding
Generation
Lpass
Lt2i
Li2t
Lfail
MME
MMBench
MMMU
MMStar
MMVP
GenEval
DPG
-
-
-
-
2283
86.1
59.0
65.1
58.0
85.8
86.1
✗
✓
✓
✓
2299 (+16)
86.6 (+0.5)
61.0 (+2.0)
65.3 (+0.2)
60.0 (+2.0)
89.4 (+3.6)
87.2 (+1.1)
✓
✗
✓
✓
2294 (+11)
87.4 (+1.3)
60.1 (+1.1)
64.9 (-0.2)
59.3 (+1.3)
88.0 (+2.2)
87.0 (+0.9)
✓
✓
✗
✓
2299 (+16)
86.6 (+0.5)
60.1 (+1.1)
65.6 (+0.5)
56.7 (-1.3)
88.8 (+3.0)
87.3 (+1.2)
✓
✓
✓
✗
2310 (+27)
86.8 (+0.7)
61.3 (+2.3)
65.3 (+0.2)
60.0 (+2.0)
88.7 (+2.9)
87.1 (+1.0)
Table 5: Loss ablations on SenseNova-U1. Dashes denote the base model without post-training. Crosses denote replacement with Linterleave on the corresponding unmodified rollouts.
Setting
GenEval
DPG
MME
MMB
MMMU
MMStar
MMVP
Base Model
85.8
86.1
2282.6
86.1
59.0
65.1
58.0
Base w/ Self-Training
87.7
86.6
2276.3
85.5
60.0
64.7
58.0
Prism w/o Rewriting
88.4
86.5
2300.6
86.8
61.0
65.6
60.7
Prism w/ Random Routing
87.6
86.5
2297.5
86.5
59.9
65.5
59.3
Prism (Ours)
89.2
87.3
2311.6
87.3
60.6
66.5
62.0
Table 6: Analysis of reliability-guided target construction on SenseNova-U1.
Figure 7: Qualitative examples generated by the model trained with Prism. The figure illustrates a diverse set of outputs, including realistic photography and artistic rendering, among others.
Appendix figures & tables14 assets
Supplementary material from the paper’s appendix.
Appendix
Rank
Model
FDT
w/o Tie
w/ Tie (0)
w/ Tie (.5)
Decisive / Total
1
SenseNova-U1 + Prism
84.73%
86.61%
77.09%
82.59%
478 / 537
2
SenseNova-U1
82.33%
83.06%
71.91%
78.62%
490 / 566
3
GPT-4o+DALL-E3
77.78%
81.38%
69.79%
76.91%
494 / 576
4
Human
76.88%
82.92%
67.50%
76.80%
486 / 597
5
Gemini1.5+Flux
62.19%
64.16%
50.36%
61.11%
438 / 558
6
BAGEL + Prism
54.38%
54.79%
42.93%
53.76%
438 / 559
Appendix
Table 8: Full OpenING arena leaderboard. We report pairwise win rates under four tie-handling protocols. The final column reports decisive battles over total battles for each entry.
Model
Prism
OneIG-Bench
Alignment
Text
Style
Reasoning
Diversity
SenseNova-U1
✗
0.873
0.942
0.405
0.326
0.170
✓
0.869
0.937
0.410
0.330
0.183
BAGEL
✗
0.813
0.366
0.395
0.189
0.176
✓
0.808
0.339
0.395
0.189
0.179
Appendix
Table 9: OneIG-Bench results. Each model generates four images per prompt on 1,120 English prompts. The five dimensions are scored with dedicated scorers and the benchmark provides no aggregate score. The better value is in bold.
Model
Prism
TIIF-Bench
Editing
Short
Long
GEdit-S
GEdit-Q
ImgEdit
SenseNova-U1
✗
0.807
0.805
7.37
6.79
3.58
✓
0.812
0.794
7.23
6.52
3.47
BAGEL
✗
0.699
0.699
6.89
6.93
3.33
✓
0.664
0.681
6.82
6.87
3.38
Appendix
Table 10: TIIF-Bench and editing results. TIIF-Bench uses a short and a long prompt variant per prompt with Qwen2.5-VL-72B as the judge. ImgEdit-Bench scores 1–5 and GEdit-Bench (English) uses VIEScore on a 10-point scale. The better value is in bold.
Action
Reliability state
Count
Share
Pass
both chains reliable
1,250
52.2%
t2i
text reliable, image flawed
571
23.8%
i2t
text flawed, image reliable
223
9.3%
Fail
both chains flawed
351
14.7%
Appendix
Table 11: Routing decisions for the first rollout of each of the 2,395 SenseNova-U1 prompt seeds.
Stage
Updates
Routing Distribution
Judge Pass Rate
Pass
t2i
i2t
Fail
Text
Image
S1
0–7
49.3%
21.3%
11.5%
17.9%
70.6%
60.8%
S2
8–15
51.6%
25.1%
8.7%
14.6%
76.7%
60.3%
S3
16–23
55.7%
25.2%
7.7%
11.4%
80.9%
63.4%
S4
24–30
57.6%
24.7%
7.1%
10.6%
82.3%
64.7%
S5
31–37
59.3%
24.2%
6.8%
9.7%
83.5%
66.1%
Appendix
Table 12: Routing dynamics across the 38 updates of one epoch on SenseNova-U1. Statistics include all judged rollouts, including replacements generated after failed rewrite verification. The judge is frozen throughout. Text and image pass rates are computed from the corresponding routing categories. All values are percentages.
Rewrite
Candidates
Accepted
Discarded
Verification Calls
t2i
571
376 (65.8%)
195 (34.2%)
1,474 ( I+ reviews)
i2t
223
117 (52.5%)
106 (47.5%)
320 ( T+ reviews)
Appendix
Table 13: Verification outcomes for rewriting candidates arising from the first-rollout cohort in Table 11 . Accepted and discarded counts are candidate-level terminal outcomes; verification calls include retries for these candidates.
Judge
Chain
Human Agreement
Judge Quality
TP
FP
TN
FN
Acc.
F1
FDR
SenseNova-U1 (frozen)
Text
113
18
47
22
80.0%
85.0%
13.7%
Image
95
17
64
24
79.5%
82.3%
15.2%
All
208
35
111
46
79.8%
83.7%
14.4%
Qwen3.7 Plus
Text
128
24
41
7
84.5%
89.2%
15.8%
Image
87
7
74
32
80.5%
81.7%
7.4%
Appendix
Table 14: Chain-level agreement with human annotations on 200 rollout trajectories. Each text or image row contains 200 judgments; the All rows aggregate 400 judgments. FDR is FP/(TP+FP) .
Judge
MME
MMBench
MMMU MC
MMStar
POPE
MMVP
GenEval
DPG
Base
2282.6
86.1
59.0
65.1
84.0
58.0
85.8
86.1
SenseNova-U1
2311.6
87.3
60.6
66.5
86.6
62.0
89.2
87.3
Qwen3.7 Plus
2304.1
87.4
60.7
66.5
86.3
59.3
88.9
86.9
Appendix
Table 15: Effects of internal and external judges on SenseNova-U1. POPE reports F1.
Figure 8: Different judgment emphases of Qwen3.7 Plus and SenseNova-U1. Qwen3.7 Plus identifies a local omission, while SenseNova-U1 identifies an image-count violation in the complete trajectory.
Judge Setting
Updates
Routing Distribution
Judge Pass Rate
Pass
t2i
i2t
Fail
Text
Image
Frozen judge (Ours)
0–7
49.3
21.3
11.5
17.9
70.6
60.8
8–15
51.6
25.1
8.7
14.6
76.7
60.3
16–23
55.7
25.2
7.7
11.4
80.9
63.4
24–30
57.6
24.7
7.1
10.6
82.3
64.7
31–37
59.3
24.2
6.8
9.7
83.5
66.1
Appendix
Table 16: Routing distributions and judge pass rates across training stages under frozen and jointly updated judges. Text and image pass rates are computed from the corresponding routing categories. All values are percentages. The jointly updated run terminates after 34 updates.
Judge Setting
Initial Zero-Image Rate
Final Zero-Image Rate
Frozen judge (Ours)
0.00%
0.00%
Jointly updated judge
0.00%
70.59%
Appendix
Table 17: Effect of freezing the judge on image-generation stability during online training. Lower zero-image rates are better.
Figure 9: Additional VQA examples. Prism corrects errors made by the base model on questions requiring visual reasoning, chart understanding, correspondence matching, spatial perception, and counting.
Figure 10: Additional generation examples on BAGEL. The generated images cover diverse subjects, scenes, and visual styles.
Figure 11: Additional interleaved generation examples. Compared with the base model, Prism produces clearer visual progression in a picture-book story, more diverse images for food recommendations, and visual understanding for a solar eclipse.
Recently, remarkable progress has been made in Unified Multimodal Models (UMMs), which integrate vision-language generation and understanding capabilities within a single framework. However, a model's strong visual understanding often fails to transfer to visual generation: it may correctly judge prompt-image alignment while failing to generate a faithful image from the same prompt. This raises a compelling question: Can a model improve itself by using its understanding module to reward its generation module? We introduce SRUM, a self-rewarding post-training framework directly applicable to existing UMMs of various designs. SRUM creates a feedback loop where the model's own understanding module acts as an internal ``evaluator'', providing corrective signals to improve generation without additional human-labeled data or external reward models. To provide comprehensive feedback, SRUM uses a global-local dual reward system: a global reward ensures overall visual semantics and layout, while a local reward refines fine-grained, object-level fidelity. SRUM shows strong generalization, boosting performance on T2I-CompBench from 82.18 to 88.37 and on T2I-ReasonBench from 43.82 to 46.75. Overall, our work establishes a powerful paradigm for enabling a UMM's understanding module to guide and enhance its own generation via self-rewarding.
Weiyang Jin, Yuwei Niu, Jiaqi Liao +4
HKU MMLab · The University of Hong Kong · Shenzhen Loop Area Institute +2
Unified Multimodal Models (UMMs) integrate image understanding and generation within a single architecture, yet how the two tasks interact remains understudied. We investigate transferability in UMMs: whether training a capability on one task improves the same capability on the other without explicit supervision. Through controlled experiments, we empirically find that transferability depends on architecture-models with fully shared transformer backbone and a unified visual encoder exhibit consistent cross-task transfer, while loosely coupled designs show little or none. Leveraging this transferability, we propose a practical training strategy. The most straightforward way to improve a target generative capability (e.g., counting) is to fine-tune generation directly, but this can degrade visual quality due to distribution shift. Instead, we train the corresponding understanding task and let it transfer into generation, which improves capability-specific generative performance while minimizing distribution shift. We validate this across three capabilities-counting, spatial relation, and text recognition/generation-showing that cross-task transferability can be systematically exploited in UMMs.
Jiwon Kang, Heeji Yoon, Jaewoo Jung +5
KAIST AI, South Korea · Blynx, South Korea · Trillionlabs, South Korea
Unified multimodal models (UMMs) understand and generate both text and images, which lets a model produce its own training data. Existing self-improvement in UMMs keeps supervision on the visual side, where image understanding judges image generation. We propose recursive cross-capability self-improvement (RSI), a training loop in which the text and visual abilities of a UMM supply training data for one another. In each round, the model generates images and reads them to find where it falls short. It then writes programs aimed at these shortcomings, and execution verifies every result against its specification. Verified renders train image generation, while labeled renders and the model's own correct programs train visual understanding and program writing. Program execution thus acts as a source of truth outside the model, so errors do not accumulate across rounds. We study RSI on charts and build BasicChartBench to evaluate open models early in training. On requests worded differently from training, four rounds of RSI raise the score from 45.7% to 60.2%, while continued training stays at 46.3%. Verified construction carries most of the gain, and targeting the model's failures adds 3.5%. Along the way, the share of verified programs rises from 48.9% to 95.2%, and the reader's accuracy on edited renders rises from 55.6% to 87.4%.
Huijuan Wang, Chufan Shi, Cheng Yang +3
University of Southern California · University of California San Diego · Carnegie Mellon University