{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/census/softmax-entropy","entry":"softmax_entropy","source":"Syntology differential census (groundwork/55, run v2_2026-09-22), per sample; not an archive number","census_date":"2026-09-22","battery_sha256":["b986f7e04d794a0d88ad4c5f32cf63ec3590b5deff0192150737bc5f1c0b4677"],"runner_sha256":["5a452d0e7c0da5b80771d1be2afe3572e253568cf5c5d59a0d08ebd663d00808"],"bucket_key":"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","bucket_fields":["rank","kind","dtype"],"claim":"Implementations sharing this entry name were each run on one shared input fixed by the positional (rank, kind, dtype) of their array arguments (the bucket). A cluster is the set whose recorded output digest (sha256 of the output rounded to 6 decimals) is identical. Identical values to six decimals on the shared input are agreement on those inputs, not a statement about the whole domain and not a substitution claim.","n_implementations_compared":18,"n_papers":29,"n_buckets":2,"n_distinct_outputs":4,"n_class_bearing":0,"not_compared":{"not_run_on_shared_input":{"n":0,"by_error":{}},"no_array_argument_ran_on_own_fixture_arguments_only":{"n":0,"n_papers":0,"recorded_shared_digest_equals_own_fixture_digest":0},"output_not_digested_non_numeric":{"n":0,"by_type":{}}},"withdrawn_excluded":0,"code_page":"/code/softmax-entropy","buckets":[{"bucket":[[2,"float","float32"]],"n_implementations_compared":15,"n_papers":25,"n_not_digested":0,"not_digested_by_type":{},"clusters":[{"output_sha":"e0770880d670f9ad","size":12,"n_papers":21,"shape":[4],"dtype":"float32","type":"Tensor","finite":true,"max_difference_among_recorded_values":0.0,"members_with_recorded_values":12,"torch_reference_conventions_with_this_digest":["tent"],"values":[1.4792883396148682,1.4573825597763062,1.8222516775131226,1.7319810390472412],"values_recorded":4,"members":[{"code_sha256_prefix":"3078648c75e3e1e3","path":"map_nav_src/models/FSTTA.py","papers":["2311.13209","2403.07366","2105.08714","2408.13983","2310.18562","2601.16240"],"paper_pages":[{"arxiv_id":"2311.13209","page":"/paper/test-time-adaptive-vision-and-language"},{"arxiv_id":"2403.07366","page":"/paper/entropy-is-not-enough-for-test-time"},{"arxiv_id":"2105.08714","page":"/paper/fighting-gradients-with-gradients-dynamic"},{"arxiv_id":"2408.13983","page":"/paper/dual-path-adversarial-lifting-for-domain"},{"arxiv_id":"2310.18562","page":"/paper/optimization-free-test-time-adaptation-for"},{"arxiv_id":"2601.16240","page":"/paper/arxiv-2601-16240"}],"arg_sig_recorded":[["x",2,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"c2df94ca3de67eb0","path":"src/main_task_wise_adamerging.py","papers":["2310.02575","2410.15430","2408.07362","2412.12153"],"paper_pages":[{"arxiv_id":"2310.02575","page":"/paper/adamerging-adaptive-model-merging-for-multi"},{"arxiv_id":"2410.15430","page":"/paper/boostadapter-improving-test-time-adaptation"},{"arxiv_id":"2408.07362","page":"/paper/badmerging-backdoor-attacks-against-model"},{"arxiv_id":"2412.12153","page":"/paper/revisiting-weight-averaging-for-model-merging"}],"arg_sig_recorded":[["x",2,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"e6171e29ed623a12","path":"domainbed/scripts/unsupervised_adaptation.py","papers":["2007.01434","2304.12566","2207.10792"],"paper_pages":[{"arxiv_id":"2007.01434","page":"/paper/in-search-of-lost-domain-generalization"},{"arxiv_id":"2304.12566","page":"/paper/adanpc-exploring-non-parametric-classifier"},{"arxiv_id":"2207.10792","page":"/paper/test-time-adaptation-via-self-training-with"}],"arg_sig_recorded":[["x",2,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"60e9cb8fea03027c","path":"src/learn_few_shots.py","papers":["2407.02880","2105.08714"],"paper_pages":[{"arxiv_id":"2407.02880","page":"/paper/knowledge-composition-using-task-vectors-with"},{"arxiv_id":"2105.08714","page":"/paper/fighting-gradients-with-gradients-dynamic"}],"arg_sig_recorded":[["x",2,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"c549267af6d34f9a","path":"TTAs-2/E-BATS.py","papers":["2506.07078","2601.16240"],"paper_pages":[{"arxiv_id":"2506.07078","page":"/paper/e-bats-efficient-backpropagation-free-test"},{"arxiv_id":"2601.16240","page":"/paper/arxiv-2601-16240"}],"arg_sig_recorded":[["x",2,"float","float32"]],"scalar_args":{"dim":"-1"},"class_bearing":false},{"code_sha256_prefix":"1bdb21f20be90a8f","path":"tta_library/zotta.py","papers":["2603.14254"],"paper_pages":[{"arxiv_id":"2603.14254","page":"/paper/arxiv-2603-14254"}],"arg_sig_recorded":[["x",2,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"1ddd5315f5202ec9","path":"eata.py","papers":["2204.02610"],"paper_pages":[{"arxiv_id":"2204.02610","page":"/paper/efficient-test-time-model-adaptation-without"}],"arg_sig_recorded":[["x",2,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"311413feee2ab1dd","path":"src/sumi.py","papers":["2503.02616"],"paper_pages":[{"arxiv_id":"2503.02616","page":"/paper/smoothing-the-shift-towards-stable-test-time"}],"arg_sig_recorded":[["x",2,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"5f3e7b3e8f7ad589","path":"tent.py","papers":["2006.10726"],"paper_pages":[{"arxiv_id":"2006.10726","page":"/paper/fully-test-time-adaptation-by-entropy"}],"arg_sig_recorded":[["x",2,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"8f4830ebee73f4a1","path":"sar.py","papers":["2302.12400"],"paper_pages":[{"arxiv_id":"2302.12400","page":"/paper/towards-stable-test-time-adaptation-in"}],"arg_sig_recorded":[["x",2,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"99c9d70a0d3706d9","path":"tta_library/beta.py","papers":["2604.15609"],"paper_pages":[{"arxiv_id":"2604.15609","page":"/paper/arxiv-2604-15609"}],"arg_sig_recorded":[["logits",2,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"be510b4c74ce8440","path":"tent.py","papers":["2006.10726"],"paper_pages":[{"arxiv_id":"2006.10726","page":"/paper/fully-test-time-adaptation-by-entropy"}],"arg_sig_recorded":[["x",2,"float","float32"]],"scalar_args":{},"class_bearing":false}]},{"output_sha":"2ce2016ced5fd7cc","size":2,"n_papers":3,"shape":[4],"dtype":"float32","type":"Tensor","finite":true,"max_difference_among_recorded_values":0.0,"members_with_recorded_values":2,"torch_reference_conventions_with_this_digest":[],"values":[1.4792802333831787,1.4573745727539062,1.8222436904907227,1.7319731712341309],"values_recorded":4,"members":[{"code_sha256_prefix":"a9edd84f4eedeec8","path":"learner/sotta.py","papers":["2310.10074","2404.01351"],"paper_pages":[{"arxiv_id":"2310.10074","page":"/paper/sotta-robust-test-time-adaptation-on-noisy-1"},{"arxiv_id":"2404.01351","page":"/paper/aetta-label-free-accuracy-estimation-for-test"}],"arg_sig_recorded":[["x",2,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"9ccab176a6977060","path":"ProtoPFormer/proto_tta.py","papers":["2604.15494"],"paper_pages":[{"arxiv_id":"2604.15494","page":"/paper/arxiv-2604-15494"}],"arg_sig_recorded":[["logits",2,"float","float32"]],"scalar_args":{},"class_bearing":false}]},{"output_sha":"7e5824d63f9fbed5","size":1,"n_papers":1,"shape":[],"dtype":"float32","type":"Tensor","finite":true,"max_difference_among_recorded_values":null,"members_with_recorded_values":1,"torch_reference_conventions_with_this_digest":[],"values":[1.6227259635925293],"values_recorded":1,"members":[{"code_sha256_prefix":"db8b634a36fc3530","path":"methods/das.py","papers":["2503.20354"],"paper_pages":[{"arxiv_id":"2503.20354","page":"/paper/surgeon-memory-adaptive-fully-test-time"}],"arg_sig_recorded":[["x",2,"float","float32"]],"scalar_args":{},"class_bearing":false}]}]},{"bucket":[[2,"float","float32"],[2,"float","float32"]],"n_implementations_compared":3,"n_papers":4,"n_not_digested":0,"not_digested_by_type":{},"clusters":[{"output_sha":"e0770880d670f9ad","size":3,"n_papers":4,"shape":[4],"dtype":"float32","type":"Tensor","finite":true,"max_difference_among_recorded_values":0.0,"members_with_recorded_values":3,"torch_reference_conventions_with_this_digest":[],"values":[1.4792883396148682,1.4573825597763062,1.8222516775131226,1.7319810390472412],"values_recorded":4,"members":[{"code_sha256_prefix":"8cc4004512e3fa2f","path":"src/adapter/petta.py","papers":["2311.18193","2309.14949"],"paper_pages":[{"arxiv_id":"2311.18193","page":"/paper/persistent-test-time-adaptation-in-episodic"},{"arxiv_id":"2309.14949","page":"/paper/towards-real-world-test-time-adaptation-tri"}],"arg_sig_recorded":[["x",2,"float","float32"],["x_ema",2,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"6aa4e7aea80769c7","path":"cotta_vit_opd.py","papers":["2605.17743"],"paper_pages":[{"arxiv_id":"2605.17743","page":"/paper/arxiv-2605-17743"}],"arg_sig_recorded":[["x",2,"float","float32"],["x_ema",2,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"867d3ea552eca91a","path":"imagenet/cotta.py","papers":["2203.13591"],"paper_pages":[{"arxiv_id":"2203.13591","page":"/paper/continual-test-time-domain-adaptation"}],"arg_sig_recorded":[["x",2,"float","float32"],["x_ema",2,"float","float32"]],"scalar_args":{},"class_bearing":false}]}]}]}