{"about":{"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.","site":"https://codewithpapers.app","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","syntology":{"site":"https://syntology.ai","developers":"https://syntology.ai/developers","mcp":{"server":"https://syntology.ai/mcp","transport":"streamable-http","server_card":"https://syntology.ai/.well-known/mcp/server-card.json","auth":{"type":"trial token, no account","trial_token":"https://syntology.ai/api/oauth/trial/token","method":"POST","docs":"https://syntology.ai/developers"}},"have":"https://syntology.ai/api/graph/have?x=<method, arXiv id or title> (free, answers coverage only)","paper_base":"https://syntology.ai/paper/","atlas_base":"https://app.syntology.ai/?focus="},"machine_readable":[{"url":"https://codewithpapers.app/llms.txt","what":"the machine catalog: every machine-readable file, counted"},{"url":"https://codewithpapers.app/index/manifest.json","what":"paper-to-code index by arXiv id, with Syntology's counts"},{"url":"https://codewithpapers.app/search/manifest.json","what":"site search index (titles, authors) and its files"},{"url":"https://codewithpapers.app/download","what":"bulk files: Syntology's layer, described there"},{"url":"https://codewithpapers.app/build_manifest.json","what":"the build record: inputs, counts, exclusions, probes"}]},"url":"/task/token-reduction","name":"Token Reduction","slug":"token-reduction","description_markdown":null,"categories":[{"name":"Adversarial","url":"/area/adversarial"},{"name":"Computer Vision","url":"/area/computer-vision"}],"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","slug_source":"derived"},"counts":{"papers_tagged":78,"papers_with_code":37,"benchmarks":0,"benchmark_tables_in_archive":0,"benchmark_tables_shown":0,"benchmark_tables_withheld_as_spam":0,"benchmark_definition":"a leaderboard table with at least one row; benchmark_tables_shown also counts the zero-row tables; benchmark_tables_in_archive adds the tables withheld as spam","datasets":0,"subtasks":0,"parent_tasks":1},"benchmarks":[],"datasets":[],"subtasks":[],"parent_tasks":[{"url":"/task/image-classification","name":"Image Classification"}],"papers":{"order":"repositories listed in the archive (most first), then date (newest first), then slug; the archive holds no stars","population":"every paper the archive tags with this task (distinct papers; spam titles not listed). 'of' counts the papers listed; until 2026-09-25 it counted only the papers with a code link, and that number is of_with_code","page":1,"pages_in_order":1,"rows_per_page":100,"shown":78,"of":78,"of_with_code":37,"tagged_in_all":78,"counts":{"archive_papers_tagged":78,"with_a_code_link":37,"where_syntology_ran_a_sample":21,"not_listed_spam_title":0,"listed":78,"listed_where_code_ran":21,"where_syntology_ran_a_sample_split":{"with_a_run_with_no_instrument_failure":15,"every_run_a_failure_of_syntologys_instrument":6,"listed_with_a_run_with_no_instrument_failure":15,"listed_every_run_a_failure_of_syntologys_instrument":6,"filter":{"states":["a run with no instrument failure","any run, instrument failures included"],"default":"a run with no instrument failure","note":"on the 'only where code ran' pages the default hides, in the browser, the rows where every run was a failure of Syntology's instrument; the second state shows them again. Rows are hidden, never re-ordered; these twins list every row"}},"definition":"distinct papers the archive tags; 'where Syntology ran a sample' counts papers with at least one harvested sample that ran, which is not a correctness claim"},"items":[{"url":"/paper/learning-compact-vision-tokens-for-efficient","slug":"learning-compact-vision-tokens-for-efficient","title":"Learning Compact Vision Tokens for Efficient Large Multimodal Models","date":"2025-06-08","arxiv_id":"2506.07138","repositories_listed":1,"syntology":null},{"url":"/paper/silvr-a-simple-language-based-video-reasoning","slug":"silvr-a-simple-language-based-video-reasoning","title":"SiLVR: A Simple Language-based Video Reasoning Framework","date":"2025-05-30","arxiv_id":"2505.24869","repositories_listed":1,"syntology":null},{"url":"/paper/one-trajectory-one-token-grounded-video","slug":"one-trajectory-one-token-grounded-video","title":"One Trajectory, One Token: Grounded Video Tokenization via Panoptic Sub-object Trajectory","date":"2025-05-29","arxiv_id":"2505.23617","repositories_listed":1,"syntology":null},{"url":"/paper/flowcut-rethinking-redundancy-via-information","slug":"flowcut-rethinking-redundancy-via-information","title":"FlowCut: Rethinking Redundancy via Information Flow for Efficient Vision-Language Models","date":"2025-05-26","arxiv_id":"2505.19536","repositories_listed":1,"syntology":{"n":1,"n_ran":1,"n_constructed":0,"n_ran_checked":0,"n_instrument":1,"n_unverified":0,"n_honours":0,"n_violates":0,"n_no_contract":0,"n_pointer_only":0,"phrase":"1 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/flowcut-rethinking-redundancy-via-information#ran","syntology_url":"https://syntology.ai/paper/2505.19536","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2505.19536"}},"official":{"repos":["tungchintao/flowcut"],"state":"official (archive's flag): 1 ran","n_ran":1,"n_constructed":0,"n_ran_no_instrument_failure":0,"n_unverified":0,"ran_from_kinds":["official"]}}},{"url":"/paper/not-all-tokens-are-what-you-need-in-thinking","slug":"not-all-tokens-are-what-you-need-in-thinking","title":"Not All Tokens Are What You Need In Thinking","date":"2025-05-23","arxiv_id":"2505.17827","repositories_listed":1,"syntology":{"n":2,"n_ran":2,"n_constructed":0,"n_ran_checked":2,"n_instrument":0,"n_unverified":0,"n_honours":0,"n_violates":0,"n_no_contract":2,"n_pointer_only":0,"phrase":"2 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/not-all-tokens-are-what-you-need-in-thinking#ran","syntology_url":"https://syntology.ai/paper/2505.17827","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2505.17827"}},"official":{"repos":["faustrazor/not-all-thinking-tokens"],"state":"official (archive's flag): 2 ran","n_ran":2,"n_constructed":0,"n_ran_no_instrument_failure":2,"n_unverified":0,"ran_from_kinds":["official"]}}},{"url":"/paper/token-reduction-should-go-beyond-efficiency","slug":"token-reduction-should-go-beyond-efficiency","title":"Token Reduction Should Go Beyond Efficiency in Generative Models -- From Vision, Language to Multimodality","date":"2025-05-23","arxiv_id":"2505.18227","repositories_listed":1,"syntology":null},{"url":"/paper/crosslmm-decoupling-long-video-sequences-from","slug":"crosslmm-decoupling-long-video-sequences-from","title":"CrossLMM: Decoupling Long Video Sequences from LMMs via Dual Cross-Attention Mechanisms","date":"2025-05-22","arxiv_id":"2505.17020","repositories_listed":1,"syntology":null},{"url":"/paper/streamline-without-sacrifice-squeeze-out","slug":"streamline-without-sacrifice-squeeze-out","title":"Streamline Without Sacrifice -- Squeeze out Computation Redundancy in LMM","date":"2025-05-21","arxiv_id":"2505.15816","repositories_listed":1,"syntology":{"n":9,"n_ran":5,"n_constructed":0,"n_ran_checked":1,"n_instrument":4,"n_unverified":4,"n_honours":0,"n_violates":1,"n_no_contract":0,"n_pointer_only":4,"phrase":"5 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 1 violated, 0 with no contract checked; 4 where Syntology's instrument failed) · 4 unverified","sample_list":"/paper/streamline-without-sacrifice-squeeze-out#ran","syntology_url":"https://syntology.ai/paper/2505.15816","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2505.15816"}},"official":{"repos":["penghao-wu/proxyv"],"state":"official (archive's flag): 1 ran","n_ran":1,"n_constructed":0,"n_ran_no_instrument_failure":0,"n_unverified":4,"ran_from_kinds":["official","unlocated"]}}},{"url":"/paper/pact-pruning-and-clustering-based-token","slug":"pact-pruning-and-clustering-based-token","title":"PACT: Pruning and Clustering-Based Token Reduction for Faster Visual Language Models","date":"2025-04-11","arxiv_id":"2504.08966","repositories_listed":1,"syntology":{"n":2,"n_ran":1,"n_constructed":0,"n_ran_checked":0,"n_instrument":1,"n_unverified":1,"n_honours":0,"n_violates":0,"n_no_contract":0,"n_pointer_only":0,"phrase":"1 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 1 unverified","sample_list":"/paper/pact-pruning-and-clustering-based-token#ran","syntology_url":"https://syntology.ai/paper/2504.08966","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2504.08966"}},"official":{"repos":["orailix/pact"],"state":"official (archive's flag): 1 ran","n_ran":1,"n_constructed":0,"n_ran_no_instrument_failure":0,"n_unverified":1,"ran_from_kinds":["official"]}}},{"url":"/paper/window-token-concatenation-for-efficient","slug":"window-token-concatenation-for-efficient","title":"Window Token Concatenation for Efficient Visual Large Language Models","date":"2025-04-05","arxiv_id":"2504.04024","repositories_listed":1,"syntology":null},{"url":"/paper/faster-parameter-efficient-tuning-with-token","slug":"faster-parameter-efficient-tuning-with-token","title":"Faster Parameter-Efficient Tuning with Token Redundancy Reduction","date":"2025-03-26","arxiv_id":"2503.20282","repositories_listed":1,"syntology":null},{"url":"/paper/when-large-vision-language-model-meets-large","slug":"when-large-vision-language-model-meets-large","title":"When Large Vision-Language Model Meets Large Remote Sensing Imagery: Coarse-to-Fine Text-Guided Token Pruning","date":"2025-03-10","arxiv_id":"2503.07588","repositories_listed":1,"syntology":null},{"url":"/paper/folder-accelerating-multi-modal-large","slug":"folder-accelerating-multi-modal-large","title":"FOLDER: Accelerating Multi-modal Large Language Models with Enhanced Performance","date":"2025-01-05","arxiv_id":"2501.02430","repositories_listed":1,"syntology":null},{"url":"/paper/cached-adaptive-token-merging-dynamic-token","slug":"cached-adaptive-token-merging-dynamic-token","title":"Cached Adaptive Token Merging: Dynamic Token Reduction and Redundant Computation Elimination in Diffusion Model","date":"2025-01-01","arxiv_id":"2501.00946","repositories_listed":1,"syntology":{"n":5,"n_ran":5,"n_constructed":0,"n_ran_checked":2,"n_instrument":3,"n_unverified":0,"n_honours":1,"n_violates":1,"n_no_contract":0,"n_pointer_only":5,"phrase":"5 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 1 honoured, 1 violated, 0 with no contract checked; 3 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/cached-adaptive-token-merging-dynamic-token#ran","syntology_url":"https://syntology.ai/paper/2501.00946","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2501.00946"}},"official":{"repos":["omidiu/ca_tome"],"state":"official (archive's flag): 5 ran","n_ran":5,"n_constructed":0,"n_ran_no_instrument_failure":2,"n_unverified":0,"ran_from_kinds":["official"]}}},{"url":"/paper/rethinking-token-reduction-with-parameter","slug":"rethinking-token-reduction-with-parameter","title":"Rethinking Token Reduction with Parameter-Efficient Fine-Tuning in ViT for Pixel-Level Tasks","date":"2025-01-01","arxiv_id":null,"repositories_listed":1,"syntology":null},{"url":"/paper/cross-layer-cache-aggregation-for-token","slug":"cross-layer-cache-aggregation-for-token","title":"Cross-Layer Cache Aggregation for Token Reduction in Ultra-Fine-Grained Image Recognition","date":"2024-12-31","arxiv_id":"2501.00243","repositories_listed":1,"syntology":null},{"url":"/paper/framefusion-combining-similarity-and","slug":"framefusion-combining-similarity-and","title":"FrameFusion: Combining Similarity and Importance for Video Token Reduction on Large Visual Language Models","date":"2024-12-30","arxiv_id":"2501.01986","repositories_listed":1,"syntology":{"n":13,"n_ran":10,"n_constructed":0,"n_ran_checked":8,"n_instrument":2,"n_unverified":3,"n_honours":0,"n_violates":0,"n_no_contract":8,"n_pointer_only":1,"phrase":"10 ran (of which 0 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 0 violated, 8 with no contract checked; 2 where Syntology's instrument failed) · 3 unverified","sample_list":"/paper/framefusion-combining-similarity-and#ran","syntology_url":"https://syntology.ai/paper/2501.01986","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2501.01986"}},"official":{"repos":["thu-nics/framefusion"],"state":"official (archive's flag): 10 ran","n_ran":10,"n_constructed":0,"n_ran_no_instrument_failure":8,"n_unverified":3,"ran_from_kinds":["official"]}}},{"url":"/paper/faster-vision-mamba-is-rebuilt-in-minutes-via","slug":"faster-vision-mamba-is-rebuilt-in-minutes-via","title":"Faster Vision Mamba is Rebuilt in Minutes via Merged Token Re-training","date":"2024-12-17","arxiv_id":"2412.12496","repositories_listed":1,"syntology":null},{"url":"/paper/enhancing-multimodal-large-language-models-2","slug":"enhancing-multimodal-large-language-models-2","title":"Enhancing Multimodal Large Language Models Complex Reason via Similarity Computation","date":"2024-12-13","arxiv_id":"2412.09817","repositories_listed":1,"syntology":null},{"url":"/paper/learning-to-merge-tokens-via-decoupled","slug":"learning-to-merge-tokens-via-decoupled","title":"Learning to Merge Tokens via Decoupled Embedding for Efficient Vision Transformers","date":"2024-12-13","arxiv_id":"2412.10569","repositories_listed":1,"syntology":{"n":34,"n_ran":28,"n_constructed":1,"n_ran_checked":12,"n_instrument":16,"n_unverified":6,"n_honours":1,"n_violates":0,"n_no_contract":11,"n_pointer_only":4,"phrase":"28 ran (of which 1 constructed an object rather than computing a result; 12 with no instrument failure: 1 honoured, 0 violated, 11 with no contract checked; 16 where Syntology's instrument failed) · 6 unverified","sample_list":"/paper/learning-to-merge-tokens-via-decoupled#ran","syntology_url":"https://syntology.ai/paper/2412.10569","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2412.10569"}},"official":{"repos":["movinghoon/dtem"],"state":"official (archive's flag): 2 ran","n_ran":2,"n_constructed":1,"n_ran_no_instrument_failure":1,"n_unverified":2,"ran_from_kinds":["found_in_text","official"]}}},{"url":"/paper/token-cropr-faster-vits-for-quite-a-few-tasks","slug":"token-cropr-faster-vits-for-quite-a-few-tasks","title":"Token Cropr: Faster ViTs for Quite a Few Tasks","date":"2024-12-01","arxiv_id":"2412.00965","repositories_listed":1,"syntology":null},{"url":"/paper/inference-optimal-vlms-need-only-one-visual","slug":"inference-optimal-vlms-need-only-one-visual","title":"Inference Optimal VLMs Need Fewer Visual Tokens and More Parameters","date":"2024-11-05","arxiv_id":"2411.03312","repositories_listed":1,"syntology":{"n":2,"n_ran":1,"n_constructed":0,"n_ran_checked":0,"n_instrument":1,"n_unverified":1,"n_honours":0,"n_violates":0,"n_no_contract":0,"n_pointer_only":2,"phrase":"1 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 1 unverified","sample_list":"/paper/inference-optimal-vlms-need-only-one-visual#ran","syntology_url":"https://syntology.ai/paper/2411.03312","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2411.03312"}},"official":{"repos":["locuslab/llava-token-compression"],"state":"official: harvested, nothing ran","n_ran":0,"n_constructed":0,"n_ran_no_instrument_failure":0,"n_unverified":1,"ran_from_kinds":["unlocated"]}}},{"url":"/paper/longvu-spatiotemporal-adaptive-compression","slug":"longvu-spatiotemporal-adaptive-compression","title":"LongVU: Spatiotemporal Adaptive Compression for Long Video-Language Understanding","date":"2024-10-22","arxiv_id":"2410.17434","repositories_listed":1,"syntology":{"n":11,"n_ran":11,"n_constructed":0,"n_ran_checked":3,"n_instrument":8,"n_unverified":0,"n_honours":0,"n_violates":0,"n_no_contract":3,"n_pointer_only":2,"phrase":"11 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 0 violated, 3 with no contract checked; 8 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/longvu-spatiotemporal-adaptive-compression#ran","syntology_url":"https://syntology.ai/paper/2410.17434","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2410.17434"}},"official":{"repos":["Vision-CAIR/LongVU"],"state":"official (archive's flag): 11 ran","n_ran":11,"n_constructed":0,"n_ran_no_instrument_failure":3,"n_unverified":0,"ran_from_kinds":["official"]}}},{"url":"/paper/rethinking-token-reduction-for-state-space","slug":"rethinking-token-reduction-for-state-space","title":"Rethinking Token Reduction for State Space Models","date":"2024-10-16","arxiv_id":"2410.14725","repositories_listed":1,"syntology":{"n":2,"n_ran":2,"n_constructed":0,"n_ran_checked":0,"n_instrument":2,"n_unverified":0,"n_honours":0,"n_violates":0,"n_no_contract":0,"n_pointer_only":2,"phrase":"2 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 2 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/rethinking-token-reduction-for-state-space#ran","syntology_url":"https://syntology.ai/paper/2410.14725","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2410.14725"}},"official":{"repos":["wuyushuwys/tor_ssm"],"state":"official (archive's flag): 2 ran","n_ran":2,"n_constructed":0,"n_ran_no_instrument_failure":0,"n_unverified":0,"ran_from_kinds":["official"]}}},{"url":"/paper/fastadasp-multitask-adapted-efficient","slug":"fastadasp-multitask-adapted-efficient","title":"FastAdaSP: Multitask-Adapted Efficient Inference for Large Speech Language Model","date":"2024-10-03","arxiv_id":"2410.03007","repositories_listed":1,"syntology":{"n":3,"n_ran":2,"n_constructed":0,"n_ran_checked":1,"n_instrument":1,"n_unverified":1,"n_honours":1,"n_violates":0,"n_no_contract":0,"n_pointer_only":3,"phrase":"2 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 1 honoured, 0 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 1 unverified","sample_list":"/paper/fastadasp-multitask-adapted-efficient#ran","syntology_url":"https://syntology.ai/paper/2410.03007","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2410.03007"}},"official":{"repos":["yichen14/fastadasp"],"state":"official (archive's flag): 2 ran","n_ran":2,"n_constructed":0,"n_ran_no_instrument_failure":1,"n_unverified":1,"ran_from_kinds":["official"]}}},{"url":"/paper/discovering-the-gems-in-early-layers","slug":"discovering-the-gems-in-early-layers","title":"Discovering the Gems in Early Layers: Accelerating Long-Context LLMs with 1000x Input Token Reduction","date":"2024-09-25","arxiv_id":"2409.17422","repositories_listed":1,"syntology":{"n":13,"n_ran":13,"n_constructed":0,"n_ran_checked":10,"n_instrument":3,"n_unverified":0,"n_honours":1,"n_violates":1,"n_no_contract":8,"n_pointer_only":3,"phrase":"13 ran (of which 0 constructed an object rather than computing a result; 10 with no instrument failure: 1 honoured, 1 violated, 8 with no contract checked; 3 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/discovering-the-gems-in-early-layers#ran","syntology_url":"https://syntology.ai/paper/2409.17422","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2409.17422"}},"official":{"repos":["salesforceairesearch/gemfilter"],"state":"official (archive's flag): 13 ran","n_ran":13,"n_constructed":0,"n_ran_no_instrument_failure":10,"n_unverified":0,"ran_from_kinds":["official"]}}},{"url":"/paper/less-is-more-a-simple-yet-effective-token","slug":"less-is-more-a-simple-yet-effective-token","title":"Less is More: A Simple yet Effective Token Reduction Method for Efficient Multi-modal LLMs","date":"2024-09-17","arxiv_id":"2409.10994","repositories_listed":1,"syntology":{"n":8,"n_ran":8,"n_constructed":0,"n_ran_checked":2,"n_instrument":6,"n_unverified":0,"n_honours":1,"n_violates":1,"n_no_contract":0,"n_pointer_only":1,"phrase":"8 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 1 honoured, 1 violated, 0 with no contract checked; 6 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/less-is-more-a-simple-yet-effective-token#ran","syntology_url":"https://syntology.ai/paper/2409.10994","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2409.10994"}},"official":{"repos":["freedomintelligence/trim"],"state":"official (archive's flag): 8 ran","n_ran":8,"n_constructed":0,"n_ran_no_instrument_failure":2,"n_unverified":0,"ran_from_kinds":["official"]}}},{"url":"/paper/bridging-local-details-and-global-context-in","slug":"bridging-local-details-and-global-context-in","title":"Bridging Local Details and Global Context in Text-Attributed Graphs","date":"2024-06-18","arxiv_id":"2406.12608","repositories_listed":1,"syntology":{"n":1,"n_ran":1,"n_constructed":1,"n_ran_checked":1,"n_instrument":0,"n_unverified":0,"n_honours":0,"n_violates":0,"n_no_contract":1,"n_pointer_only":1,"phrase":"1 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified; the one sample that ran constructed an object rather than computing a result","sample_list":"/paper/bridging-local-details-and-global-context-in#ran","syntology_url":"https://syntology.ai/paper/2406.12608","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2406.12608"}},"official":{"repos":["wykk00/graphbridge"],"state":"official (archive's flag): 1 ran","n_ran":1,"n_constructed":1,"n_ran_no_instrument_failure":1,"n_unverified":0,"ran_from_kinds":["official"]}}},{"url":"/paper/algm-adaptive-local-then-global-token-merging-1","slug":"algm-adaptive-local-then-global-token-merging-1","title":"ALGM: Adaptive Local-then-Global Token Merging for Efficient Semantic Segmentation with Plain Vision Transformers","date":"2024-06-14","arxiv_id":"2406.09936","repositories_listed":1,"syntology":{"n":12,"n_ran":7,"n_constructed":0,"n_ran_checked":6,"n_instrument":1,"n_unverified":5,"n_honours":3,"n_violates":0,"n_no_contract":3,"n_pointer_only":12,"phrase":"7 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 3 honoured, 0 violated, 3 with no contract checked; 1 where Syntology's instrument failed) · 5 unverified","sample_list":"/paper/algm-adaptive-local-then-global-token-merging-1#ran","syntology_url":"https://syntology.ai/paper/2406.09936","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2406.09936"}},"official":{"repos":["tue-mps/algm-segmenter"],"state":"official (archive's flag): 7 ran","n_ran":7,"n_constructed":0,"n_ran_no_instrument_failure":6,"n_unverified":5,"ran_from_kinds":["official"]}}},{"url":"/paper/hierarchical-context-merging-better-long","slug":"hierarchical-context-merging-better-long","title":"Hierarchical Context Merging: Better Long Context Understanding for Pre-trained LLMs","date":"2024-04-16","arxiv_id":"2404.10308","repositories_listed":1,"syntology":{"n":15,"n_ran":12,"n_constructed":1,"n_ran_checked":7,"n_instrument":5,"n_unverified":3,"n_honours":1,"n_violates":0,"n_no_contract":6,"n_pointer_only":15,"phrase":"12 ran (of which 1 constructed an object rather than computing a result; 7 with no instrument failure: 1 honoured, 0 violated, 6 with no contract checked; 5 where Syntology's instrument failed) · 3 unverified","sample_list":"/paper/hierarchical-context-merging-better-long#ran","syntology_url":"https://syntology.ai/paper/2404.10308","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2404.10308"}},"official":{"repos":["alinlab/homer"],"state":"official (archive's flag): 12 ran","n_ran":12,"n_constructed":1,"n_ran_no_instrument_failure":7,"n_unverified":3,"ran_from_kinds":["official"]}}},{"url":"/paper/llava-prumerge-adaptive-token-reduction-for","slug":"llava-prumerge-adaptive-token-reduction-for","title":"LLaVA-PruMerge: Adaptive Token Reduction for Efficient Large Multimodal Models","date":"2024-03-22","arxiv_id":"2403.15388","repositories_listed":1,"syntology":{"n":2,"n_ran":2,"n_constructed":0,"n_ran_checked":0,"n_instrument":2,"n_unverified":0,"n_honours":0,"n_violates":0,"n_no_contract":0,"n_pointer_only":0,"phrase":"2 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 2 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/llava-prumerge-adaptive-token-reduction-for#ran","syntology_url":"https://syntology.ai/paper/2403.15388","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2403.15388"}},"official":null}},{"url":"/paper/haltingvt-adaptive-token-halting-transformer","slug":"haltingvt-adaptive-token-halting-transformer","title":"HaltingVT: Adaptive Token Halting Transformer for Efficient Video Recognition","date":"2024-01-10","arxiv_id":"2401.04975","repositories_listed":1,"syntology":null},{"url":"/paper/which-tokens-to-use-investigating-token","slug":"which-tokens-to-use-investigating-token","title":"Which Tokens to Use? Investigating Token Reduction in Vision Transformers","date":"2023-08-09","arxiv_id":"2308.04657","repositories_listed":1,"syntology":{"n":8,"n_ran":8,"n_constructed":0,"n_ran_checked":3,"n_instrument":5,"n_unverified":0,"n_honours":0,"n_violates":0,"n_no_contract":3,"n_pointer_only":3,"phrase":"8 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 0 violated, 3 with no contract checked; 5 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/which-tokens-to-use-investigating-token#ran","syntology_url":"https://syntology.ai/paper/2308.04657","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2308.04657"}},"official":null}},{"url":"/paper/content-aware-token-sharing-for-efficient-1","slug":"content-aware-token-sharing-for-efficient-1","title":"Content-aware Token Sharing for Efficient Semantic Segmentation with Vision Transformers","date":"2023-06-03","arxiv_id":"2306.02095","repositories_listed":1,"syntology":null},{"url":"/paper/pumer-pruning-and-merging-tokens-for","slug":"pumer-pruning-and-merging-tokens-for","title":"PuMer: Pruning and Merging Tokens for Efficient Vision Language Models","date":"2023-05-27","arxiv_id":"2305.17530","repositories_listed":1,"syntology":{"n":3,"n_ran":3,"n_constructed":2,"n_ran_checked":2,"n_instrument":1,"n_unverified":0,"n_honours":0,"n_violates":0,"n_no_contract":2,"n_pointer_only":0,"phrase":"3 ran (of which 2 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/pumer-pruning-and-merging-tokens-for#ran","syntology_url":"https://syntology.ai/paper/2305.17530","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2305.17530"}},"official":{"repos":["csarron/pumer"],"state":"official (archive's flag): 3 ran","n_ran":3,"n_constructed":2,"n_ran_no_instrument_failure":2,"n_unverified":0,"ran_from_kinds":["official"]}}},{"url":"/paper/adavit-adaptive-tokens-for-efficient-vision","slug":"adavit-adaptive-tokens-for-efficient-vision","title":"AdaViT: Adaptive Tokens for Efficient Vision Transformer","date":"2021-12-14","arxiv_id":"2112.07658","repositories_listed":1,"syntology":{"n":7,"n_ran":6,"n_constructed":0,"n_ran_checked":3,"n_instrument":3,"n_unverified":1,"n_honours":0,"n_violates":0,"n_no_contract":3,"n_pointer_only":7,"phrase":"6 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 0 violated, 3 with no contract checked; 3 where Syntology's instrument failed) · 1 unverified","sample_list":"/paper/adavit-adaptive-tokens-for-efficient-vision#ran","syntology_url":"https://syntology.ai/paper/2112.07658","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2112.07658"}},"official":null}},{"url":"/paper/tr-bert-dynamic-token-reduction-for","slug":"tr-bert-dynamic-token-reduction-for","title":"TR-BERT: Dynamic Token Reduction for Accelerating BERT Inference","date":"2021-05-25","arxiv_id":"2105.11618","repositories_listed":1,"syntology":null},{"url":null,"slug":"token-transforming-a-unified-and-training","title":"Token Transforming: A Unified and Training-Free Token Compression Framework for Vision Transformer Acceleration","date":"2025-06-06","arxiv_id":"2506.05709","repositories_listed":0,"syntology":null},{"url":null,"slug":"astraea-a-gpu-oriented-token-wise","title":"Astraea: A GPU-Oriented Token-wise Acceleration Framework for Video Diffusion Transformers","date":"2025-06-05","arxiv_id":"2506.05096","repositories_listed":0,"syntology":null},{"url":null,"slug":"towards-storage-efficient-visual-document","title":"Towards Storage-Efficient Visual Document Retrieval: An Empirical Study on Reducing Patch-Level Embeddings","date":"2025-06-05","arxiv_id":"2506.04997","repositories_listed":0,"syntology":null},{"url":null,"slug":"vscan-rethinking-visual-token-reduction-for","title":"VScan: Rethinking Visual Token Reduction for Efficient Large Vision-Language Models","date":"2025-05-28","arxiv_id":"2505.22654","repositories_listed":0,"syntology":null},{"url":null,"slug":"the-overthinker-s-diet-cutting-token-calories","title":"The Overthinker's DIET: Cutting Token Calories with DIfficulty-AwarE Training","date":"2025-05-25","arxiv_id":"2505.19217","repositories_listed":0,"syntology":null},{"url":null,"slug":"plan-and-budget-effective-and-efficient-test","title":"Plan and Budget: Effective and Efficient Test-Time Scaling on Large Language Model Reasoning","date":"2025-05-22","arxiv_id":"2505.16122","repositories_listed":0,"syntology":null},{"url":null,"slug":"drp-distilled-reasoning-pruning-with-skill","title":"DRP: Distilled Reasoning Pruning with Skill-aware Step Decomposition for Efficient Large Reasoning Models","date":"2025-05-20","arxiv_id":"2505.13975","repositories_listed":0,"syntology":null},{"url":null,"slug":"star-stage-wise-attention-guided-token","title":"STAR: Stage-Wise Attention-Guided Token Reduction for Efficient Large Vision-Language Models Inference","date":"2025-05-18","arxiv_id":"2505.12359","repositories_listed":0,"syntology":null},{"url":"/paper/attend-to-not-attended-structure-then-detail","slug":"attend-to-not-attended-structure-then-detail","title":"Attend to Not Attended: Structure-then-Detail Token Merging for Post-training DiT Acceleration","date":"2025-05-16","arxiv_id":"2505.11707","repositories_listed":0,"syntology":{"n":1,"n_ran":1,"n_constructed":0,"n_ran_checked":0,"n_instrument":1,"n_unverified":0,"n_honours":0,"n_violates":0,"n_no_contract":0,"n_pointer_only":1,"phrase":"1 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/attend-to-not-attended-structure-then-detail#ran","syntology_url":"https://syntology.ai/paper/2505.11707","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2505.11707"}},"official":null}},{"url":null,"slug":"ecosaferag-efficient-security-through-context","title":"EcoSafeRAG: Efficient Security through Context Analysis in Retrieval-Augmented Generation","date":"2025-05-16","arxiv_id":"2505.13506","repositories_listed":0,"syntology":null},{"url":null,"slug":"hypernym-mercury-token-optimization-through","title":"Hypernym Mercury: Token Optimization Through Semantic Field Constriction And Reconstruction From Hypernyms. A New Text Compression Method","date":"2025-05-12","arxiv_id":"2505.08058","repositories_listed":0,"syntology":null},{"url":null,"slug":"dymu-dynamic-merging-and-virtual-unmerging","title":"DyMU: Dynamic Merging and Virtual Unmerging for Efficient VLMs","date":"2025-04-23","arxiv_id":"2504.17040","repositories_listed":0,"syntology":null},{"url":null,"slug":"zipr1-reinforcing-token-sparsity-in-mllms","title":"ZipR1: Reinforcing Token Sparsity in MLLMs","date":"2025-04-23","arxiv_id":"2504.18579","repositories_listed":0,"syntology":null},{"url":null,"slug":"dynamic-compressing-prompts-for-efficient","title":"Dynamic Compressing Prompts for Efficient Inference of Large Language Models","date":"2025-04-15","arxiv_id":"2504.11004","repositories_listed":0,"syntology":null},{"url":null,"slug":"efficient-llama-3-2-vision-by-trimming-cross","title":"Efficient LLaMA-3.2-Vision by Trimming Cross-attended Visual Features","date":"2025-04-01","arxiv_id":"2504.00557","repositories_listed":0,"syntology":null},{"url":null,"slug":"local-information-matters-inference","title":"Local Information Matters: Inference Acceleration For Grounded Conversation Generation Models Through Adaptive Local-Aware Token Pruning","date":"2025-03-31","arxiv_id":"2503.23959","repositories_listed":0,"syntology":null},{"url":null,"slug":"token-dynamics-towards-efficient-and-dynamic","title":"Token Dynamics: Towards Efficient and Dynamic Video Token Representation for Video Large Language Models","date":"2025-03-21","arxiv_id":"2503.16980","repositories_listed":0,"syntology":null},{"url":null,"slug":"vamba-understanding-hour-long-videos-with","title":"Vamba: Understanding Hour-Long Videos with Hybrid Mamba-Transformers","date":"2025-03-14","arxiv_id":"2503.11579","repositories_listed":0,"syntology":null},{"url":null,"slug":"layton-latent-consistency-tokenizer-for-1024","title":"Layton: Latent Consistency Tokenizer for 1024-pixel Image Reconstruction and Generation by 256 Tokens","date":"2025-03-11","arxiv_id":"2503.08377","repositories_listed":0,"syntology":null},{"url":null,"slug":"silent-hazards-of-token-reduction-in-vision","title":"Does Acceleration Cause Hidden Instability in Vision Language Models? Uncovering Instance-Level Divergence Through a Large-Scale Empirical Study","date":"2025-03-09","arxiv_id":"2503.06794","repositories_listed":0,"syntology":null},{"url":null,"slug":"batchgemba-token-efficient-machine","title":"BatchGEMBA: Token-Efficient Machine Translation Evaluation with Batched Prompting and Prompt Compression","date":"2025-03-04","arxiv_id":"2503.02756","repositories_listed":0,"syntology":null},{"url":null,"slug":"knowing-when-to-stop-dynamic-context-cutoff","title":"Knowing When to Stop: Dynamic Context Cutoff for Large Language Models","date":"2025-02-03","arxiv_id":"2502.01025","repositories_listed":0,"syntology":null},{"url":null,"slug":"mint-mitigating-hallucinations-in-large","title":"MINT: Mitigating Hallucinations in Large Vision-Language Models via Token Reduction","date":"2025-02-02","arxiv_id":"2502.00717","repositories_listed":0,"syntology":null},{"url":null,"slug":"learning-free-token-reduction-for-multi-modal","title":"Learning Free Token Reduction for Multi-Modal Large Language Models","date":"2025-01-29","arxiv_id":"2501.17391","repositories_listed":0,"syntology":null},{"url":null,"slug":"dynamic-token-reduction-during-generation-for","title":"Dynamic Token Reduction during Generation for Vision Language Models","date":"2025-01-24","arxiv_id":"2501.14204","repositories_listed":0,"syntology":null},{"url":null,"slug":"adafv-accelerating-vlms-with-self-adaptive","title":"AdaFV: Rethinking of Visual-Language alignment for VLM acceleration","date":"2025-01-16","arxiv_id":"2501.09532","repositories_listed":0,"syntology":null},{"url":null,"slug":"imagepiece-content-aware-re-tokenization-for","title":"ImagePiece: Content-aware Re-tokenization for Efficient Image Recognition","date":"2024-12-21","arxiv_id":"2412.16491","repositories_listed":0,"syntology":null},{"url":null,"slug":"deploying-foundation-model-powered-agent","title":"Deploying Foundation Model Powered Agent Services: A Survey","date":"2024-12-18","arxiv_id":"2412.13437","repositories_listed":0,"syntology":null},{"url":null,"slug":"asymrnr-video-diffusion-transformers","title":"AsymRnR: Video Diffusion Transformers Acceleration with Asymmetric Reduction and Restoration","date":"2024-12-16","arxiv_id":"2412.11706","repositories_listed":0,"syntology":null},{"url":null,"slug":"trim-token-reduction-and-inference-modeling","title":"TRIM: Token Reduction and Inference Modeling for Cost-Effective Language Generation","date":"2024-12-10","arxiv_id":"2412.07682","repositories_listed":0,"syntology":null},{"url":null,"slug":"accelerating-multimodal-large-language-models-1","title":"Accelerating Multimodal Large Language Models by Searching Optimal Vision Token Reduction","date":"2024-11-30","arxiv_id":"2412.00556","repositories_listed":0,"syntology":null},{"url":null,"slug":"efficient-multi-modal-large-language-models","title":"Efficient Multi-modal Large Language Models via Visual Token Grouping","date":"2024-11-26","arxiv_id":"2411.17773","repositories_listed":0,"syntology":null},{"url":null,"slug":"rethinking-token-reduction-in-mllms-towards-a","title":"Rethinking Token Reduction in MLLMs: Towards a Unified Paradigm for Training-Free Acceleration","date":"2024-11-26","arxiv_id":"2411.17686","repositories_listed":0,"syntology":null},{"url":null,"slug":"freepruner-a-training-free-approach-for-large","title":"freePruner: A Training-free Approach for Large Multimodal Model Acceleration","date":"2024-11-23","arxiv_id":"2411.15446","repositories_listed":0,"syntology":null},{"url":null,"slug":"retrieval-replace-reduction-an-effective","title":"PAR: Prompt-Aware Token Reduction Method for Efficient Large Multimodal Models","date":"2024-10-09","arxiv_id":"2410.07278","repositories_listed":0,"syntology":null},{"url":null,"slug":"cut-the-crap-an-economical-communication","title":"Cut the Crap: An Economical Communication Pipeline for LLM-based Multi-Agent Systems","date":"2024-10-03","arxiv_id":"2410.02506","repositories_listed":0,"syntology":null},{"url":null,"slug":"vote-mix-plug-and-play-token-reduction-for","title":"Vote&Mix: Plug-and-Play Token Reduction for Efficient Vision Transformer","date":"2024-08-30","arxiv_id":"2408.17062","repositories_listed":0,"syntology":null},{"url":null,"slug":"fit-rag-black-box-rag-with-factual","title":"FIT-RAG: Black-Box RAG with Factual Information and Token Reduction","date":"2024-03-21","arxiv_id":"2403.14374","repositories_listed":0,"syntology":null},{"url":null,"slug":"token-propagation-controller-for-efficient","title":"TPC-ViT: Token Propagation Controller for Efficient Vision Transformer","date":"2024-01-03","arxiv_id":"2401.01470","repositories_listed":0,"syntology":null},{"url":"/paper/selective-structured-state-spaces-for-long","slug":"selective-structured-state-spaces-for-long","title":"Selective Structured State-Spaces for Long-Form Video Understanding","date":"2023-03-25","arxiv_id":"2303.14526","repositories_listed":0,"syntology":null},{"url":null,"slug":"tore-token-reduction-for-efficient-human-mesh","title":"TORE: Token Reduction for Efficient Human Mesh Recovery with Transformer","date":"2022-11-19","arxiv_id":"2211.10705","repositories_listed":0,"syntology":null}],"syntology_records":21,"syntology_note":"a paper without a record is not a recorded non-run: it may lack an arXiv id or simply be absent from the graph layer"},"key_notes":{"n_ran_checked":"legacy name, kept unchanged so existing readers do not break: it counts the samples that ran with no instrument failure (honoured, violated, and ran with no contract checked); it does not mean a contract was checked, and the pages print it as 'K with no instrument failure', not 'K checked'","n_constructed":"a sub-count of the samples that ran, never subtracted from them and never a failure: an executed sample whose run returned an instance of its own class (fixture_out_type equals the entry name): the run built an object and did not compute a result (Syntology's RAN record, counts.constructed)"},"since_archive_judged":{"papers":0,"papers_placed":0,"judged_by":{"model":"a model","models_as_recorded":[],"from":"lb_rows_run_verified.jsonl: each record's `model` field (the extractor's call that wrote decisions[])"},"note":"a model's judgement from the paper's own tables that they report on this leaderboard, not a result: no value from a paper that was not placed is printed or carried here; papers counts each paper once over these leaderboards","backfill":"8,886 of 18,637 newer papers checked so far (733 more had nothing to place; 41 could not be read)","backfill_complete":false,"leaderboards":[]},"papers_where_code_ran":21,"papers_where_code_ran_first_page":"/task/token-reduction/papers/ran/1","papers_where_code_ran_caption":"We ran code from the paper's repository; we did not run it on this task or check it against the task's benchmarks.","pages":[{"order":"ran","page":1,"of":1,"url":"/task/token-reduction/papers/ran/1","json":"/task/token-reduction/papers/ran/1.json","rows":21}],"description_links":{"kept":0,"unwrapped_to_text":0,"bare_urls_linked":0,"relative_images_dropped":0,"rule":"internal links are kept only when the target slug exists in the catalog"},"syntology":{"read_at":"2026-10-05T00:36:11+00:00","claim":"Per-sample execution status on synthesized fixtures ('ran N of M samples'); not a correctness claim and not a ranking signal.","status_vocabulary":{"ran_honours":"ran, honoured the contract we drafted","ran_violates":"ran, violated the contract we drafted","ran_draft_wrong":"ran; our contract draft was wrong, not the code","ran_fixture":"ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)","ran":"ran on a synthesized input","unverified":"unverified (harvested, no recorded run)"},"row_route":"each syntology object on a benchmark's first row and on a paper carries sample_list (this site's paper page, #ran), syntology_url (syntology.ai's page for the paper, only when that route probed live) and mcp (the agent call that lists the paper's harvested samples; not a verdict on the paper)","developers":"https://syntology.ai/developers"},"not_shown":{"libraries":"the archive has no per-task library table","trend_sparklines":"the Trend column of the benchmarks table was a rendered image; it is not in the archive","social_and_latest_sorts":"stars and social signals are not in the archive"},"record_sha256":"f98656f4ef6b8eb1f97d4897da044abb56a035fba2008c01afc406ecba5bb3c1","record_changed_at":"2026-10-05","record_changed_at_basis":"changed"}