Home › Census › get_1d_sincos_pos_embed_from_grid

get_1d_sincos_pos_embed_from_grid

Syntologyfunction-name censuscensus 2026-09-22battery b986f7e04d79all samples with this name

get_1d_sincos_pos_embed_from_grid: 8 implementations from 17 papers ran on one shared input (census 2026-09-22, battery b986f7e04d79); they produced 2 distinct outputs across 2 buckets, one shared input per bucket.

Identical values to six decimals (the recorded digest) on the shared input are agreement on those inputs, not equivalence. Implementations are compared only within one bucket, the positional (rank, kind, dtype) of each array argument; the argument name is not part of the key because the harness draws the shared array from (rank, kind) and casts it to the dtype, whatever the name; each bucket's shared input is fixed by that key, so members of one bucket saw bitwise-identical inputs under their own scalar arguments. A cluster is the set of members whose recorded output digest is identical. Nothing here says which computation a paper's method intended, and nothing reproduces a paper's results.

Not compared, and not in the tables or the counts above:

Every examined implementation of this name that took an array argument ran on the shared input.
1 more implementation from 1 paper took no array argument and ran only on its own fixture arguments; no shared input existed for it, so it is not compared (on 1 the recorded shared-output digest is the own-fixture digest).
Every compared output was digested.

Bucket 1 of 2: arg 1: rank 1, kind float, dtype float64

6 implementations from 13 papers share this bucket (rank, kind, dtype of each array argument, positional; each member's recorded signature, argument name included, is shown under it); 1 distinct output, largest cluster first. Values are the first 8 of the recorded output, flattened.

Cluster (same digest to six decimals)	Members	Shared output on this bucket's input
6 implementations 13 papers f7de1e939f56 recorded values differ by up to 2.39e-09	Scalable Diffusion Models with Transformers PixArt-$α$: Fast Training of Diffusion Transformer for... PixArt-Σ: Weak-to-Strong Training of Diffusion... Causal Diffusion Transformers for Generative Modeling Robust-U1: Can MLLMs Self-Recover Corrupted Visual... added by Syntology and 1 more paper holding this code (all in the JSON twin) one code sha held from 6 papers' repositories models_dig.py e5947aba recorded pos:1/float/float64; embed_dim=8 CroCo: Self-Supervised Pre-training for 3D Vision Tasks... arXiv:2506.21803 Meta-Learning Hyperparameters for Parameter Efficient Fine-Tuning added by Syntology one code sha held from 3 papers' repositories models/croco.py 4bdd36ea recorded pos:1/float/float64; embed_dim=8 PACT: Self-Evolving Physical Safety Alignment for... added by Syntology models/rdt/model.py 06a688b1 recorded pos:1/float/float64; embed_dim=8 Diffusion Models as Masked Autoencoders models_cross.py 6a53acdc recorded pos:1/float/float64; embed_dim=8, idx_range=14 UDT: Reconciling U-Nets and Diffusion Transformers with... added by Syntology models/UDT.py 7b75bf9e recorded pos:1/float/float64; embed_dim=8 EchoJEPA: A Latent Predictive Foundation Model for... added by Syntology src/models/predictor.py f6dd6f91 recorded pos:1/float/float64; embed_dim=8	[-0.785738, -0.0902659, -0.00903877, -0.000903889, 0.618559, 0.995918, 0.999959, 1, …] shape [8, 8] · float64 · ndarray

Bucket 2 of 2: arg 1: rank 1, kind float, dtype float32

2 implementations from 4 papers share this bucket (rank, kind, dtype of each array argument, positional; each member's recorded signature, argument name included, is shown under it); 1 distinct output, largest cluster first. Values are the first 8 of the recorded output, flattened.

Cluster (same digest to six decimals)	Members	Shared output on this bucket's input
2 implementations 4 papers f7de1e939f56 recorded values differ by up to 5.96e-08	An Image is Worth 16x16 Words: Transformers for Image... UrbanDiT: A Foundation Model for Open-World Urban... DreamVLA: A Vision-Language-Action Model Dreamed with... one code sha held from 3 papers' repositories models/dreamvla_model.py 5c8ea92d recorded pos:1/float/float32; embed_dim=8 Self-supervised Learning of Echocardiographic Video... models/modeling_pretrain.py 44aacaf6 recorded pos:1/float/float32; embed_dim=8	[-0.785738, -0.0902659, -0.00903877, -0.000903889, 0.618559, 0.995918, 0.999959, 1, …] shape [8, 8] · float32 · Tensor

Identical values to six decimals (the recorded digest) on the shared input are agreement on those inputs, not equivalence; where a cluster's members carry recorded values, the largest difference among them is shown under the cluster. Paper titles are the archive's archive 2025-07-28 where the paper is in the archive and the graph's where it was added by Syntology; papers with no page here are shown by their recorded paper id only. A paper count above the implementation count means one implementation (one code sha) is held from several papers' repositories and counts once. Per-sample status, licence and fingerprint records for each paper are on its paper page. JSON twin: /census/get-1d-sincos-pos-embed-from-grid.json.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections