- Rating
- 3
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 3
- Soundness
- 2
Read review
Paper summary
This paper presents COFFEE, this is a method for plug-and-play inference time guidance for discrete diffusion models which works with denoiser token predictions and a target independent probabilistic carrier to model dependencies between tokens. This method works by compiling sequence level constraints and rewards into finite-state representations which allows efficient dynamic programming to account for future completions and then samples jointly guided sequences without retraining the diffusion model. This method is evaluated with symbolic reasoning, language generation, and biological sequence design, against a comprehensive set of baselines.
Critical strengths
- COFFEE demonstrates strong empirical results across a diverse range of domains and tasks, including symbolic reasoning, language generation, and biological sequence design. This is important for demonstrating broad applicability.
- The paper provides extensive comparisons against a broad set of existing guidance methods, including recent approaches. These comparisons strengthen the empirical evidence supporting COFFEE's effectiveness relative to prior work.
- The evaluation on multiple pretrained diffusion backbones, including LLaDA and Dream, provides evidence of the method's transferability demonstrating its not limited to a particular model architecture.
- COFFEE can be integrated with pretrained diffusion models without modifying or retraining their parameters. This makes the approach attractive for applying new constraints and objectives to existing models.
Critical weaknesses
-
A of COFFEE is the trade-off between objective optimization and generation diversity. Specifically, on DeepSTARR, the method produces only a single unique sequence, while on K562, only 32 distinct sequences are generated among 1,000 samples. This could limit the practical usefulness of the method for biological sequence design. It is worth noting that this limitation is explicitly acknowledged by the authors and does not directly negate the otherwise convincing empirical evidence presented in the paper, but would benefit from additional experimentation (See questions for authors).
-
COFFEE presents a useful framework, several of its core ideas are very directly built upon existing approaches. For example:
-
Ctrl-G: Combined probabilistic sequence models with finite-state automata and dynamic programming for constrained generation, closely resembling COFFEE's use of paired states and backward inference to guide future completions.
-
CDD: Introduced inference-time constraint enforcement through projection during discrete diffusion denoising, establishing an earlier approach to steering diffusion trajectories toward satisfying constraints.
-
CoDD: Introduced structured probabilistic modeling of token dependencies in discrete diffusion, which COFFEE directly builds upon through its target-independent dependence carrier.
COFFEE does well in effectively integrates joint dependency modeling with compiled objectives, and demonstrates strong empirical results. However, its methodological novelty appears somewhat incremental, as the core techniques are established in prior work.
Questions for authors
-
It would be helpful for the authors further clarify the main methodological advances of COFFEE relative to existing approaches such as CoDD and Ctrl-G. For example, would be helpful for the authors to make more clear what capabilities does COFFEE introduce more than combining established dependency modeling and finite-state inference techniques.
-
Given the significant diversity collapse observed on DeepSTARR and K562, the authors would benefit from much deeper investigation of the trade-off between objective optimization and generation diversity. These results would be very helpful.
-
A few of the experiments, such as Sudoku and Dyck repair, involve substantial task-specific construction of feasible solution spaces. Could the authors further isolate the benefits of COFFEE's joint guidance mechanism from those of the underlying feasibility construction. An additional ablation of this would be informative.
6 comments
W2 and Q1. Why COFFEE is defined by the scattered-mask query and advances both communities (Revised Secs. 2.3 and 3.5; App. J, Table 23)
COFFEE is defined by the scattered-mask inference query rather than an additive pairing of existing parts. Combining an efficient density model and an efficient constraint scorer does not automatically make their expectation tractable under arbitrary query patterns (Khosravi et al., 2019), yet parallel discrete diffusion requires evaluating E_q₀[W_C(X₀) | x_t, X₀ⁱ = u] across scattered masked positions M_t where multiple non-adjacent tokens commit simultaneously. Answering this query in a single O(L · A · |V| · H²) sweep requires aligning carrier transitions and objective states on a shared product graph sᵢ = (hᵢ, aᵢ) (Eq. 13), advancing both communities in distinct ways.
1. Neurosymbolic and probabilistic-circuit advances (Ctrl-G, TRACE, CoDD; App. J, Table 23). Autoregressive neurosymbolic controllers (Ctrl-G, TRACE) never face scattered parallel masks because unresolved tokens form a contiguous suffix with one known boundary; additionally, Ctrl-G handles only hard deterministic text constraints, while TRACE uses local token-linear weights that miss multi-token biological motifs (nZVs Q2). COFFEE unifies both via weighted regular-language Aho-Corasick automata supporting hard and multi-token soft constraints over parallel diffusion masks. Conversely, whereas CoDD and Discrete Copula Diffusion model unconditional token dependencies without a control objective, COFFEE stacks a modular objective graph over a CoDD-style dependency carrier (App. C.1, Table 11), where the carrier acts as a data prior curbing reward exploitation during parallel commits.
2. Discrete-diffusion guidance advances (CDD, DG-TAG). Continuous simplex projections (CDD) and predictor gradients (DG-TAG) update relaxed per-position marginals without tracking discrete automaton states, which leads to hard grammar violations in repair (Table 3), enhancer mode collapse (Table 2), and homopolymer reward hacking on 3IXT protein scaffolding (Table 35). By running exact backward-forward DP over discrete states, COFFEE coordinates parallel commits and composes finite-state diversity rules (HR16/HR24, B_t) in one pass without per-step backpropagation.
| Method family | Decoding topology | Supported objectives | Parallel coordination & diversity invariants |
|---|---|---|---|
| Ctrl-G / TRACE | Left-to-right AR suffix | Hard DFA only (Ctrl-G) or token-linear (TRACE) | Sequential single-token commits; no parallel diffusion |
| CoDD / Discrete Copula | Parallel discrete diffusion | None (unconditional dependency modeling only) | Captures joint over M_t; no reward or constraint layer |
| CDD / DG-TAG | Parallel discrete diffusion | Continuous simplex projection or predictor gradients | Relaxed marginals; breaks hard repair & repeat bounds |
| COFFEE (Ours) | Parallel scattered masks M_t | Unified hard DFA + multi-token soft regular weights | Exact product-state DP + composable HR16/HR24 & B_t |
Table note. Detailed capability matrix across all baseline families appears in Appendix J (Table 23).
Q2 and W1. Optimization-diversity trade-off and inference-time remedies (New Sec. 3.3, Table 2; App. K, Tables 24 to 28)
Evaluating Diversity = N_unique / N over official 249-bp (DeepSTARR) and 200-bp (K562) windows shows that COFFEE maintains near-ceiling uniqueness across ten of twelve main settings without extra penalties (App. M.2, Table 36), confirming that diversity loss is not a general side effect of DP lookahead. Instead, concentration arises strictly in unconstrained DeepSTARR and K562 from two separable factors, namely neural enhancer predictors assigning extreme scores to poly-T repeats and cooled replay (τₐ = 0.3) sharpening onto those peaks. Because Eq. 13 acts over discrete states, two inference-time automata target distinct layers of diversity without retraining. HR16 and HR24 enforce within-sequence homopolymer limits, while the exclusion bank (B_t) prevents exact duplication of previously delivered evaluator windows.
1. Separating replay temperature from oracle reward bias (App. M.2, Table 36). Restoring neutral temperature (τₐ = 1) leaves our ten diverse tasks steady while lifting decoded uniqueness markedly on K562; by contrast, DeepSTARR remains concentrated because its oracle peaks on poly-T tracts even at unit temperature.
2. Composable homopolymer bounds (HR16, HR24) and batch deduplication (B_t, Sec. 3.3, Table 2; App. K). Composing 16-bp or 24-bp repeat bounds with six-mer enhancer weights (GATA-1 and HNF4A motifs in Table 41, using a validation-calibrated soft scale on K562) prevents poly-T shortcuts during backward message passing and redirects score mass onto diverse regulatory motifs above competing baselines at natural genomic perplexity. Independently, freezing a prefix-tree exclusion bank (B_t) over delivered evaluator windows avoids exact cross-sample duplicates in N=32 and growing N=128 batches (accelerated cleanly via vectorized state projection), and the two automata compose directly when both repeat caps and zero batch duplicates are desired.
| Diagnostic or N=1,000 method | DS Target% | DS Valid Hits | DS LOO | K562 Target% | K562 Valid Hits | K562 LOO |
|---|---|---|---|---|---|---|
| Other 10 main settings (Table 36) | N/A | N/A | N/A | N/A | 97.5% to 100% unique | N/A |
| CDD / D-CBG (HR16 valid hits) | 100.0% / 5.1% | 159 / 51 | 1.20 / 1799.7 | 42.5% / 100.0% | 423 / 1 | 70.1 / 1.20 |
| DG-TAG / MDM-VGB (HR16 valid hits) | 99.2% / 60.8% | 441 / 597 | 2.60 / 56.6 | 97.2% / 84.7% | 818 / 835 | 3.40 / 55.5 |
| COFFEE soft only (τₐ = 0.3, HR16) | 100.0% (all-T 5.31) | 0 (1 uniq) | 1.05 | 99.5% | 5 (28 uniq) | 1.08 |
| COFFEE soft only (τₐ = 1, uniq /1000) | 100.0% (score 5.28) | 20 uniq | 1.06 | 85.1% | 638 uniq | 1.84 |
| HR16 hard DFA only (W_soft ≡ 1) | 13.9% | 139 (174 uniq) | 1.32 | 43.3% | 357 (914 uniq) | 2.99 |
| COFFEE + HR16 (Table 2) | 96.5% | 819 (841 uniq) | 2.73 | 99.9% | 846 (847 uniq) | 3.93 |
| COFFEE + HR24 (App. K) | 99.3% | 946 (950 uniq) | 2.91 | 99.9% | 833 (MDM-VGB 845) | 4.00 |
| Duplicate-exclusion bank (B_t) | DS Unique | DS Target% | DS LOO | K562 Unique | K562 Target% | K562 Speedup |
|---|---|---|---|---|---|---|
| Empty (N=32) to Growing (N=128) | 1/32 to 128/128 | 100% / 99.2% | 1.05 / 1.22 | 2/32 to 128/128 | 100% / 96.1% | 8.58× (LOO 1.30) |
Table note. DS denotes DeepSTARR (249-bp window, target 1.17) and K562 uses 200-bp views; full 7-baseline HR16/HR24 and bank ablations appear in Sec. 3.3 (Table 2) and App. K (Tables 24 to 28).
We thank the reviewer for recognizing our empirical results across diverse domains, broad baseline comparisons, and plug-and-play transferability across pretrained diffusion backbones (
LLaDA,Dream,D3LM, andEvoDiff). Your questions directly shaped Sections 3.3 and 3.4 and our clarified methodological positioning in Section 3.5 and Appendix J (highlighted in sapphire blue in the revised PDF).