Home › Census › softmax_entropy

softmax_entropy

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

softmax_entropy: 18 implementations from 29 papers ran on one shared input (census 2026-09-22, battery b986f7e04d79); they produced 4 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.
No implementation of this name ran on its own fixture arguments only: every one that ran took an array argument.
Every compared output was digested.
The harness also ran torch's own computation of this name under 3 named conventions; those rows are not implementations, are not counted, and only label a cluster whose digest they share.

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

15 implementations from 25 papers share this bucket (rank, kind, dtype of each array argument, positional; each member's recorded signature, argument name included, is shown under it); 3 distinct outputs, 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
12 implementations 21 papers e0770880d670 recorded values identical same digest as torch: tent	Fast-Slow Test-Time Adaptation for Online... Entropy is not Enough for Test-Time Adaptation: From the... Fighting Gradients with Gradients: Dynamic Defenses... Dual-Path Adversarial Lifting for Domain Shift... Optimization-Free Test-Time Adaptation for Cross-Person... and 1 more paper holding this code (all in the JSON twin) one code sha held from 6 papers' repositories map_nav_src/models/FSTTA.py 3078648c recorded x:2/float/float32 AdaMerging: Adaptive Model Merging for Multi-Task Learning BoostAdapter: Improving Vision-Language Test-Time... BadMerging: Backdoor Attacks Against Model Merging Revisiting Weight Averaging for Model Merging one code sha held from 4 papers' repositories src/main_task_wise_adamerging.py c2df94ca recorded x:2/float/float32 In Search of Lost Domain Generalization AdaNPC: Exploring Non-Parametric Classifier for... Test-Time Adaptation via Self-Training with Nearest... one code sha held from 3 papers' repositories domainbed/scripts/unsupervised_adaptation.py e6171e29 recorded x:2/float/float32 Knowledge Composition using Task Vectors with Learned... Fighting Gradients with Gradients: Dynamic Defenses... one code sha held from 2 papers' repositories src/learn_few_shots.py 60e9cb8f recorded x:2/float/float32 E-BATS: Efficient Backpropagation-Free Test-Time... Test-Time Adaptation for Speech Emotion Recognition added by Syntology one code sha held from 2 papers' repositories TTAs-2/E-BATS.py c549267a recorded x:2/float/float32; dim=-1 ZOTTA: Test-Time Adaptation with Gradient-Free... added by Syntology tta_library/zotta.py 1bdb21f2 recorded x:2/float/float32 Efficient Test-Time Model Adaptation without Forgetting eata.py 1ddd5315 recorded x:2/float/float32 Smoothing the Shift: Towards Stable Test-Time Adaptation... src/sumi.py 311413fe recorded x:2/float/float32 Tent: Fully Test-time Adaptation by Entropy Minimization tent.py 5f3e7b3e recorded x:2/float/float32 Towards Stable Test-Time Adaptation in Dynamic Wild World sar.py 8f4830eb recorded x:2/float/float32 Adapting in the Dark: Efficient and Stable Test-Time... added by Syntology tta_library/beta.py 99c9d70a recorded logits:2/float/float32 Tent: Fully Test-time Adaptation by Entropy Minimization tent.py be510b4c recorded x:2/float/float32	[1.47929, 1.45738, 1.82225, 1.73198] shape [4] · float32 · Tensor
2 implementations 3 papers 2ce2016ced5f recorded values identical	SoTTA: Robust Test-Time Adaptation on Noisy Data Streams AETTA: Label-Free Accuracy Estimation for Test-Time Adaptation one code sha held from 2 papers' repositories learner/sotta.py a9edd84f recorded x:2/float/float32 ICLR 2026 Test-Time Updates (TTU) Workshop PROTOTTA:... added by Syntology ProtoPFormer/proto_tta.py 9ccab176 recorded logits:2/float/float32	[1.47928, 1.45737, 1.82224, 1.73197] shape [4] · float32 · Tensor
1 implementation 1 paper 7e5824d63f9f	SURGEON: Memory-Adaptive Fully Test-Time Adaptation via... methods/das.py db8b634a recorded x:2/float/float32	[1.62273] shape [] · float32 · Tensor

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

3 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
3 implementations 4 papers e0770880d670 recorded values identical	Persistent Test-time Adaptation in Recurring Testing Scenarios Towards Real-World Test-Time Adaptation: Tri-Net... one code sha held from 2 papers' repositories src/adapter/petta.py 8cc40045 recorded x:2/float/float32, x_ema:2/float/float32 MoASE++: Mixture of Activation Sparsity Experts with... added by Syntology cotta_vit_opd.py 6aa4e7ae recorded x:2/float/float32, x_ema:2/float/float32 Continual Test-Time Domain Adaptation imagenet/cotta.py 867d3ea5 recorded x:2/float/float32, x_ema:2/float/float32	[1.47929, 1.45738, 1.82225, 1.73198] shape [4] · 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/softmax-entropy.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