{"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/cosine-similarity","entry":"cosine_similarity","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":26,"n_papers":30,"n_buckets":5,"n_distinct_outputs":11,"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":9,"n_papers":10,"recorded_shared_digest_equals_own_fixture_digest":9},"output_not_digested_non_numeric":{"n":0,"by_type":{}}},"withdrawn_excluded":0,"code_page":"/code/cosine-similarity","buckets":[{"bucket":[[2,"float","float32"],[2,"float","float32"]],"n_implementations_compared":13,"n_papers":16,"n_not_digested":0,"not_digested_by_type":{},"clusters":[{"output_sha":"adc71adfbba8ca30","size":5,"n_papers":5,"shape":[4,4],"dtype":"float32","type":"Tensor","finite":true,"max_difference_among_recorded_values":0.0,"members_with_recorded_values":5,"torch_reference_conventions_with_this_digest":["matrix"],"values":[0.9977779388427734,-0.11758613586425781,0.7382721900939941,-0.2374095916748047,-0.11758613586425781,0.9974908828735352,-0.398309588432312,0.19110107421875,0.7382721900939941,-0.398309588432312,1.0007596015930176,-0.31519222259521484,-0.2374095916748047,0.19110107421875,-0.31519222259521484,0.9986743927001953],"values_recorded":16,"members":[{"code_sha256_prefix":"0b4f1ab07f515c0d","path":"model.py","papers":["2602.10622"],"paper_pages":[{"arxiv_id":"2602.10622","page":"/paper/arxiv-2602-10622"}],"arg_sig_recorded":[["x",2,"float","float32"],["y",2,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"19bb073cdac57b77","path":"modeling.py","papers":["2602.14492"],"paper_pages":[{"arxiv_id":"2602.14492","page":"/paper/arxiv-2602-14492"}],"arg_sig_recorded":[["x",2,"float","float32"],["y",2,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"3ae294e827bbb579","path":"models/proto/protaugment.py","papers":["2105.12995"],"paper_pages":[{"arxiv_id":"2105.12995","page":"/paper/protaugment-unsupervised-diverse-short-texts"}],"arg_sig_recorded":[["x",2,"float","float32"],["y",2,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"810e3d44d48121de","path":"dora/dora.py","papers":["2206.04530"],"paper_pages":[{"arxiv_id":"2206.04530","page":"/paper/dora-exploring-outlier-representations-in"}],"arg_sig_recorded":[["a",2,"float","float32"],["b",2,"float","float32"]],"scalar_args":{"eps":"1e-05"},"class_bearing":false},{"code_sha256_prefix":"88695493222917ca","path":"laft/laft.py","papers":["2503.01184"],"paper_pages":[{"arxiv_id":"2503.01184","page":"/paper/language-assisted-feature-transformation-for"}],"arg_sig_recorded":[["x1",2,"float","float32"],["x2",2,"float","float32"]],"scalar_args":{"eps":"1e-08"},"class_bearing":false}]},{"output_sha":"c914e8188e43fff1","size":4,"n_papers":6,"shape":[4],"dtype":"float32","type":"Tensor","finite":true,"max_difference_among_recorded_values":0.0,"members_with_recorded_values":4,"torch_reference_conventions_with_this_digest":["rowwise"],"values":[1.0,1.0,1.0000001192092896,0.9999999403953552],"values_recorded":4,"members":[{"code_sha256_prefix":"2a673044a613aa62","path":"train_watermark_model.py","papers":["2310.06356","2402.14007"],"paper_pages":[{"arxiv_id":"2310.06356","page":"/paper/a-semantic-invariant-robust-watermark-for"},{"arxiv_id":"2402.14007","page":"/paper/can-watermarks-survive-translation-on-the"}],"arg_sig_recorded":[["x",2,"float","float32"],["y",2,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"38ede58bd9091f21","path":"segmentation/losses/dist_kd.py","papers":["2205.10536","2406.01512"],"paper_pages":[{"arxiv_id":"2205.10536","page":"/paper/knowledge-distillation-from-a-stronger"},{"arxiv_id":"2406.01512","page":"/paper/mad-multi-alignment-meg-to-text-decoding"}],"arg_sig_recorded":[["x",2,"float","float32"],["y",2,"float","float32"]],"scalar_args":{"eps":"1e-08"},"class_bearing":false},{"code_sha256_prefix":"b5f44526e6cc6a9d","path":"lib/models/losses/dist_kd.py","papers":["2205.10536","2509.23871"],"paper_pages":[{"arxiv_id":"2205.10536","page":"/paper/knowledge-distillation-from-a-stronger"},{"arxiv_id":"2509.23871","page":"/paper/arxiv-2509-23871"}],"arg_sig_recorded":[["a",2,"float","float32"],["b",2,"float","float32"]],"scalar_args":{"eps":"1e-08"},"class_bearing":false},{"code_sha256_prefix":"e0c14774d8283a63","path":"evaluation/caculate_R_precison.py","papers":["1605.05396"],"paper_pages":[{"arxiv_id":"1605.05396","page":"/paper/generative-adversarial-text-to-image"}],"arg_sig_recorded":[["x1",2,"float","float32"],["x2",2,"float","float32"]],"scalar_args":{"dim":"1","eps":"1e-08"},"class_bearing":false}]},{"output_sha":"ff59a10244815501","size":2,"n_papers":3,"shape":[4,4],"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":[1.000539779663086,-0.11911584436893463,0.739213764667511,-0.23958921432495117,-0.11911584436893463,1.0039740800857544,-0.400068461894989,0.19246912002563477,0.739213764667511,-0.400068461894989,1.0006215572357178,-0.3152473568916321,-0.23958921432495117,0.19246912002563477,-0.3152473568916321,0.9991869330406189],"values_recorded":16,"members":[{"code_sha256_prefix":"e157ae27e2ac5b4a","path":"train_dit.py","papers":["2311.15529","2606.18209"],"paper_pages":[{"arxiv_id":"2311.15529","page":"/paper/efficient-dataset-distillation-via-minimax"},{"arxiv_id":"2606.18209","page":"/paper/arxiv-2606-18209"}],"arg_sig_recorded":[["ta",2,"float","float32"],["tb",2,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"bcb8d066adb36735","path":"src/analysis/cluster_analysis.py","papers":["2501.03012"],"paper_pages":[{"arxiv_id":"2501.03012","page":"/paper/analyzing-fine-tuning-representation-shift"}],"arg_sig_recorded":[["batch1",2,"float","float32"],["batch2",2,"float","float32"]],"scalar_args":{"eps":"1e-08"},"class_bearing":false}]},{"output_sha":"13169b86a884d25f","size":1,"n_papers":1,"shape":[4,4],"dtype":"float32","type":"ndarray","finite":true,"max_difference_among_recorded_values":null,"members_with_recorded_values":1,"torch_reference_conventions_with_this_digest":[],"values":[1.0,-0.11897356063127518,0.7391290664672852,-0.23933760821819305,-0.11897356063127518,0.9999998211860657,-0.3990154564380646,0.19153061509132385,0.7391290664672852,-0.3990154564380646,1.0,-0.31492578983306885,-0.23933760821819305,0.19153061509132385,-0.31492578983306885,1.0],"values_recorded":16,"members":[{"code_sha256_prefix":"892eedee9677ae9e","path":"evaluation/score_single.py","papers":["2304.04565"],"paper_pages":[{"arxiv_id":"2304.04565","page":"/paper/soccernet-caption-dense-video-captioning-for"}],"arg_sig_recorded":[["vec1",2,"float","float32"],["vec2",2,"float","float32"]],"scalar_args":{},"class_bearing":false}]},{"output_sha":"e331ee3b84bf670e","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":[32.41746520996094],"values_recorded":1,"members":[{"code_sha256_prefix":"2be8edc2eec0bb61","path":"main_lip.py","papers":["2607.00671"],"paper_pages":[{"arxiv_id":"2607.00671","page":"/paper/arxiv-2607-00671"}],"arg_sig_recorded":[["grad1",2,"float","float32"],["grad2",2,"float","float32"]],"scalar_args":{},"class_bearing":false}]}]},{"bucket":[[1,"float","float64"],[1,"float","float64"]],"n_implementations_compared":9,"n_papers":10,"n_not_digested":0,"not_digested_by_type":{},"clusters":[{"output_sha":"6c3c396ed6b5c36d","size":7,"n_papers":8,"shape":[],"dtype":"float64","type":"float64","finite":true,"max_difference_among_recorded_values":2.220446049250313e-16,"members_with_recorded_values":7,"torch_reference_conventions_with_this_digest":[],"values":[0.9999999999999999],"values_recorded":1,"members":[{"code_sha256_prefix":"91467442d43e757d","path":"sentence_chunk.py","papers":["2311.08721","2506.00773"],"paper_pages":[{"arxiv_id":"2311.08721","page":"/paper/a-robust-semantics-based-watermark-for-large"},{"arxiv_id":"2506.00773","page":"/paper/dynamic-chunking-and-selection-for-reading"}],"arg_sig_recorded":[["vec1",1,"float","float64"],["vec2",1,"float","float64"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"07e648dea34560c3","path":"ReSo/agent_graph/agent_graph.py","papers":["2503.02390"],"paper_pages":[{"arxiv_id":"2503.02390","page":"/paper/reso-a-reward-driven-self-organizing-llm"}],"arg_sig_recorded":[["vec1",1,"float","float64"],["vec2",1,"float","float64"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"273c1c7bcd72f9da","path":"explanatory_contextual_retrieval.py","papers":["2310.19923"],"paper_pages":[{"arxiv_id":"2310.19923","page":"/paper/jina-embeddings-2-8192-token-general-purpose"}],"arg_sig_recorded":[["vector1",1,"float","float64"],["vector2",1,"float","float64"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"440007deabcc0d03","path":"scripts/experiment.py","papers":["2410.03168"],"paper_pages":[{"arxiv_id":"2410.03168","page":"/paper/can-watermarked-llms-be-identified-by-users"}],"arg_sig_recorded":[["vec_a",1,"float","float64"],["vec_b",1,"float","float64"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"729c43fab0e64a4e","path":"infer/infer_gpt.py","papers":["2410.08143"],"paper_pages":[{"arxiv_id":"2410.08143","page":"/paper/delta-an-online-document-level-translation"}],"arg_sig_recorded":[["a",1,"float","float64"],["b",1,"float","float64"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"ac76b4e960077b4a","path":"utils/metrics.py","papers":["2412.15241"],"paper_pages":[{"arxiv_id":"2412.15241","page":"/paper/quantifying-positional-biases-in-text"}],"arg_sig_recorded":[["x",1,"float","float64"],["y",1,"float","float64"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"dbe1b491ed11294e","path":"compute_mexa.py","papers":["2410.05873"],"paper_pages":[{"arxiv_id":"2410.05873","page":"/paper/mexa-multilingual-evaluation-of-english"}],"arg_sig_recorded":[["array1",1,"float","float64"],["array2",1,"float","float64"]],"scalar_args":{},"class_bearing":false}]},{"output_sha":"96a52f1bd6096c20","size":1,"n_papers":1,"shape":[],"dtype":"float64","type":"float64","finite":true,"max_difference_among_recorded_values":null,"members_with_recorded_values":1,"torch_reference_conventions_with_this_digest":[],"values":[0.9999979779921274],"values_recorded":1,"members":[{"code_sha256_prefix":"99b2c601b3157cc1","path":"optex_jax/OptEx_JAX.py","papers":["2402.11427"],"paper_pages":[{"arxiv_id":"2402.11427","page":"/paper/optex-expediting-first-order-optimization"}],"arg_sig_recorded":[["vector1",1,"float","float64"],["vector2",1,"float","float64"]],"scalar_args":{},"class_bearing":false}]},{"output_sha":"af5570f5a1810b7a","size":1,"n_papers":1,"shape":[],"dtype":"float64","type":"float64","finite":true,"max_difference_among_recorded_values":null,"members_with_recorded_values":1,"torch_reference_conventions_with_this_digest":[],"values":[0.0],"values_recorded":1,"members":[{"code_sha256_prefix":"1eb9d4562f7a1cbf","path":"poisoning/attack_eval_pipeline.py","papers":["2209.15266"],"paper_pages":[{"arxiv_id":"2209.15266","page":"/paper/data-poisoning-attacks-against-multimodal"}],"arg_sig_recorded":[["a",1,"float","float64"],["b",1,"float","float64"]],"scalar_args":{},"class_bearing":false}]}]},{"bucket":[[1,"float","float32"],[1,"float","float32"]],"n_implementations_compared":2,"n_papers":2,"n_not_digested":0,"not_digested_by_type":{},"clusters":[{"output_sha":"6c3c396ed6b5c36d","size":2,"n_papers":2,"shape":[],"dtype":"float64","type":"float","finite":true,"max_difference_among_recorded_values":0.0,"members_with_recorded_values":2,"torch_reference_conventions_with_this_digest":["matrix","rowwise"],"values":[0.9999998807907104],"values_recorded":1,"members":[{"code_sha256_prefix":"6dc29ca9372db5fe","path":"four_layer_convergence.py","papers":["2207.12316"],"paper_pages":[{"arxiv_id":"2207.12316","page":"/paper/a-theoretical-framework-for-inference-and"}],"arg_sig_recorded":[["x1",1,"float","float32"],["x2",1,"float","float32"]],"scalar_args":{},"class_bearing":false},{"code_sha256_prefix":"e8a410a4cc2f0aeb","path":"src/losses.py","papers":["2303.05506"],"paper_pages":[{"arxiv_id":"2303.05506","page":"/paper/tangos-regularizing-tabular-neural-networks"}],"arg_sig_recorded":[["w1",1,"float","float32"],["w2",1,"float","float32"]],"scalar_args":{},"class_bearing":false}]}]},{"bucket":[[3,"float","float32"],[3,"float","float32"]],"n_implementations_compared":1,"n_papers":1,"n_not_digested":0,"not_digested_by_type":{},"clusters":[{"output_sha":"d669d7c421a33d34","size":1,"n_papers":1,"shape":[2,4,4],"dtype":"float32","type":"Tensor","finite":true,"max_difference_among_recorded_values":null,"members_with_recorded_values":1,"torch_reference_conventions_with_this_digest":[],"values":[0.9999999403953552,0.4665004312992096,-0.6605871319770813,-0.07094205915927887,0.4665004312992096,1.0000001192092896,-0.6909473538398743,-0.2041459083557129,-0.6605871319770813,-0.6909473538398743,1.0,0.6115278005599976,-0.07094205915927887,-0.2041459083557129,0.6115278005599976,1.0,1.0,0.47027936577796936,-0.16421456634998322,0.24748122692108154,0.47027936577796936,1.0000001192092896,-0.0540490485727787,-0.046573154628276825,-0.16421456634998322,-0.0540490485727787,0.9999999403953552,0.7503538131713867,0.24748122692108154,-0.046573154628276825,0.7503538131713867,0.9999999403953552],"values_recorded":32,"members":[{"code_sha256_prefix":"e0ea79d7ea081390","path":"utils/proto_utils.py","papers":["2006.11325"],"paper_pages":[{"arxiv_id":"2006.11325","page":"/paper/self-supervised-prototypical-transfer"}],"arg_sig_recorded":[["x",3,"float","float32"],["y",3,"float","float32"]],"scalar_args":{},"class_bearing":false}]}]},{"bucket":[[4,"float","float32"],[4,"float","float32"]],"n_implementations_compared":1,"n_papers":1,"n_not_digested":0,"not_digested_by_type":{},"clusters":[{"output_sha":"acfc7c36fce590b1","size":1,"n_papers":1,"shape":[2,4,4],"dtype":"float32","type":"Tensor","finite":true,"max_difference_among_recorded_values":null,"members_with_recorded_values":1,"torch_reference_conventions_with_this_digest":[],"values":[0.9999999403953552,0.9999998807907104,0.9999999403953552,0.9999998807907104,0.9999999403953552,1.0,1.0,0.9999999403953552,0.9999998807907104,1.0,1.0,1.0,0.9999998211860657,0.9999999403953552,1.0,0.9999999403953552,0.9999999403953552,1.0,1.0,1.0,0.9999999403953552,1.0,1.0,1.0,1.0,1.0,1.0,1.0,0.9999999403953552,1.0,0.9999999403953552,0.9999998807907104],"values_recorded":32,"members":[{"code_sha256_prefix":"6d87036209039489","path":"models.py","papers":["1804.03189"],"paper_pages":[{"arxiv_id":"1804.03189","page":"/paper/deep-painterly-harmonization"}],"arg_sig_recorded":[["x",4,"float","float32"],["y",4,"float","float32"]],"scalar_args":{"patch_size":"3"},"class_bearing":false}]}]}]}