Prediction-Driven Top-K Jaccard Coefficient

• The paper introduces a dynamic refinement of Jaccard similarity by predicting optimal Top-K neighbors using neural models and a Sparse Differential Transformer.
• It employs a supervised prediction task to adaptively determine neighborhood sizes, which significantly enhances robustness and discriminative power in clustering.
• Empirical results on large-scale datasets demonstrate improved face clustering performance and generalization across multiple domains.

The prediction-driven Top-K Jaccard similarity coefficient is a dynamic, data-adaptive refinement of the traditional Jaccard approach for measuring pairwise relationships in face clustering graphs. Central to this methodology is the replacement of a static, globally fixed neighbor count with an individually predicted optimal Top-K for each node, determined through a supervised prediction task powered by neural models and further stabilized via a Sparse Differential Transformer (SDT). The resulting framework achieves increased discriminative power, improved robustness to noise, and state-of-the-art clustering performance on multiple large-scale datasets (Zhang et al., 27 Dec 2025).

1. Mathematical Foundations of the Top-K Jaccard Similarity

The classical Jaccard similarity for two sets $A$ and $B$ is given by:

$J(A, B) = \frac{|A \cap B|}{|A \cup B|}.$

In the context of face clustering, the sets $N_i$ and $N_j$ typically correspond to the K nearest neighbors of nodes $i$ and $j$ in the embedding space, measured, for instance, by cosine similarity:

$$J(N_i, N_j) = \frac{|N_i \cap N_j|}{|N_i \cup N_j|}.$$

In the prediction-driven Top-K extension, the neighbor count $K$ for each node is not fixed but predicted. If $\mathrm{Top}\text{-}K_i$ is the predicted number for node $B$0 (rounded to $B$1), then

$B$2

$B$3

with the intersection $B$4.

The prediction-driven Top-K Jaccard edge probability is then:

$B$5

where $B$6 is a normalized pairwise similarity, obtained via distance-to-probability transformation:

$B$7

$B$8

$B$9

This approach increases the reliability of similarity measurements by focusing on a purified, node-specific neighborhood (Zhang et al., 27 Dec 2025).

2. Data-Driven Prediction of Optimal Top-K

Rather than applying a fixed neighbor threshold, the optimal neighborhood size for each node is formalized as a supervised prediction problem. For node $J(A, B) = \frac{|A \cap B|}{|A \cup B|}.$0, the model considers the top-$J(A, B) = \frac{|A \cap B|}{|A \cup B|}.$1 candidate neighbors (ranked by cosine similarity) and predicts a score $J(A, B) = \frac{|A \cap B|}{|A \cup B|}.$2 for each candidate $J(A, B) = \frac{|A \cap B|}{|A \cup B|}.$3, approximating the likelihood that $J(A, B) = \frac{|A \cap B|}{|A \cup B|}.$4 shares the same identity as $J(A, B) = \frac{|A \cap B|}{|A \cup B|}.$5.

During inference, the predicted Top-K is obtained by thresholding:

$J(A, B) = \frac{|A \cap B|}{|A \cup B|}.$6

where $J(A, B) = \frac{|A \cap B|}{|A \cup B|}.$7 is typically set to 0.90.

The input to the predictor consists of the embedding sequence $J(A, B) = \frac{|A \cap B|}{|A \cup B|}.$8 and corresponding positional or relative distance encodings. These are projected into queries, keys, and values for Transformer-based processing.

Training employs a binary cross-entropy loss:

$J(A, B) = \frac{|A \cap B|}{|A \cup B|}.$9

with $N_i$0 denoting the ground-truth same-identity indicator. Training samples primarily target a window around the true decision boundary to maintain class balance (Zhang et al., 27 Dec 2025).

3. Transformer-Based Predictors and the Sparse Differential Transformer (SDT)

The initial prediction framework employs a vanilla Transformer architecture (three layers, eight self-attention heads, hidden size 1024) with standard self-attention. However, the vanilla Transformer tends to overvalue irrelevant relationships, introducing noise into the neighborhood prediction.

Differential Self-Attention

The attention computation is refined by splitting queries and keys, and applying a subtractive update:

$N_i$1

where $N_i$2 denotes softmax and $N_i$3 is learnable.

Sparse Differential Transformer (SDT)

Sparsity is imposed by masking all but the top-K entries per row:

$N_i$4

with $N_i$5 denoting the Top-K mask. This structure masks weak key–query pairs, focusing the model on the most informative local relationships and improving resilience to noise.

Mixture-of-Experts SDT

MoE-SDT introduces a mixture over masks $N_i$6 (with $N_i$7) and combines their outputs using learnable weights $N_i$8:

$N_i$9

This construct further mitigates prediction errors near the Top-K boundary (Zhang et al., 27 Dec 2025).

4. Integration Into Clustering Workflow

The integration of the prediction-driven Top-K Jaccard coefficient into face clustering follows these algorithmic steps:

1. Extract face embeddings $N_j$0.
2. For each $N_j$1, determine top-$N_j$2 candidates via cosine similarity.
3. Process $N_j$3 using the SDT predictor, yielding scores $N_j$4.
4. Compute $N_j$5.
5. Round to $N_j$6 (e.g., nearest multiple of 10 up to $N_j$7).
6. Build refined neighborhoods $N_j$8 and compute $N_j$9.
7. Threshold $i$0 to construct a sparse adjacency matrix.
8. Apply the Map-Equation codec for community detection.

Pseudocode for the procedure is:

$j$1 (Zhang et al., 27 Dec 2025)

5. Empirical Results and Robustness

Evaluation on several benchmarks demonstrates the efficacy of the prediction-driven Top-K Jaccard approach.

MS-Celeb-1M Clustering:

Method	$i$1 (584K)	$i$2 (584K)	$i$3 (5.21M)	$i$4 (5.21M)
FC-ESER	95.28	93.85	89.40	88.80
Diff-Cluster	95.46	94.14	90.08	89.14

Sigmoid distance-transform yields +0.18% $i$5, +0.16% $i$6 versus exponential; adding Top-K filtering yields +0.35% $i$7, +0.46% $i$8.

Transformer Variant Ablations:

Architecture	$i$9	$j$0
Vanilla Transformer	94.25	92.73
Vanilla+Top-K Mask	94.78	93.25
Differential (no mask)	95.05	93.59
Differential+Top-K (SDT)	95.34	93.93
MoE-SDT	95.46	94.14

Noisy similarity matrix experiments simulate up to 40% noise: SDT maintains pairwise F-score, while vanilla Transformer degrades significantly.

Generalization evaluations confirm improvements across MSMT17 (person re-ID) and DeepFashion, as well as gains from substituting SDT into contemporary clustering pipelines (Zhang et al., 27 Dec 2025).

6. Impact and Significance

The prediction-driven Top-K Jaccard similarity coefficient constitutes an advancement in clustering methodology by dynamically adapting local connectivity based on learned data relationships rather than static heuristics. By defining neighborhood boundaries as a supervised prediction problem and employing SDT for robust selection, the approach produces neighborhood sets with enhanced purity, resulting in more reliable graph-based similarity estimation.

Empirical evidence establishes increases in overall clustering accuracy, robustness to random noise in similarity calculations, and improved performance generalization into domains beyond face clustering. A plausible implication is that prediction-driven Top-K filtering can be fruitfully adapted to related graph construction problems in metric learning and instance-level retrieval beyond the specific setting investigated.

In summary, this method advances the state of the art in graph-based face clustering with potential for transferability across adjacent domains (Zhang et al., 27 Dec 2025).