{"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":"/paper/pretraining-text-encoders-with-adversarial-1","title":"Pretraining Text Encoders with Adversarial Mixture of Training Signal Generators","arxiv_id":"2204.03243","date":"2022-04-07","proceeding":"ICLR 2022 4","authors":["Yu Meng","Chenyan Xiong","Payal Bajaj","Saurabh Tiwary","Paul Bennett","Jiawei Han","Xia Song"],"abstract":"We present a new framework AMOS that pretrains text encoders with an Adversarial learning curriculum via a Mixture Of Signals from multiple auxiliary generators. Following ELECTRA-style pretraining, the main encoder is trained as a discriminator to detect replaced tokens generated by auxiliary masked language models (MLMs). Different from ELECTRA which trains one MLM as the generator, we jointly train multiple MLMs of different sizes to provide training signals at various levels of difficulty. To push the discriminator to learn better with challenging replaced tokens, we learn mixture weights over the auxiliary MLMs' outputs to maximize the discriminator loss by backpropagating the gradient from the discriminator via Gumbel-Softmax. For better pretraining efficiency, we propose a way to assemble multiple MLMs into one unified auxiliary model. AMOS outperforms ELECTRA and recent state-of-the-art pretrained models by about 1 point on the GLUE benchmark for BERT base-sized models.","url_abs":"https://arxiv.org/abs/2204.03243v1","url_pdf":"https://arxiv.org/pdf/2204.03243v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"pretraining-text-encoders-with-adversarial-1","repo_url":"https://github.com/microsoft/amos","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[],"methods":[{"method_slug":"adam","method_name":"Adam"},{"method_slug":"attention","method_name":"Attention"},{"method_slug":"attention-dropout","method_name":"Attention Dropout"},{"method_slug":"bert","method_name":"BERT"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"electra","method_name":"ELECTRA"},{"method_slug":"layer-normalization","method_name":"Layer Normalization"},{"method_slug":"linear-layer","method_name":"Linear Layer"},{"method_slug":"linear-warmup-with-linear-decay","method_name":"Linear Warmup With Linear Decay"},{"method_slug":"multi-head-attention","method_name":"Multi-Head Attention"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"weight-decay","method_name":"Weight Decay"},{"method_slug":"wordpiece","method_name":"WordPiece"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2204.03243","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2204.03243"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/microsoft/amos","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran_draft_wrong":1,"ran_fixture":1,"ran":1,"unverified":5},"by_repo_kind":{"official":{"samples":8,"ran":3,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"9df40357afea56cc","entry":"to_list","repo":"microsoft/amos","repo_kind":"official","path":"huggingface/run_squad.py","file_url":"https://github.com/microsoft/amos/blob/HEAD/huggingface/run_squad.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":2,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9df40357afea56cc"}},{"code_sha256_prefix":"cac113ca87b9d9f3","entry":"acc_and_f1","repo":"microsoft/amos","repo_kind":"official","path":"huggingface/utils_for_glue.py","file_url":"https://github.com/microsoft/amos/blob/HEAD/huggingface/utils_for_glue.py","link_basis":"harvester_set","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"cac113ca87b9d9f3"}},{"code_sha256_prefix":"56a2f49e4e3b9f21","entry":"relative_position_bucket","repo":"microsoft/amos","repo_kind":"official","path":"huggingface/amos/modeling_amos.py","file_url":"https://github.com/microsoft/amos/blob/HEAD/huggingface/amos/modeling_amos.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"56a2f49e4e3b9f21"}},{"code_sha256_prefix":"12cd41e0ee10b80e","entry":"get_optimizer_grouped_parameters","repo":"microsoft/amos","repo_kind":"official","path":"huggingface/run_glue.py","file_url":"https://github.com/microsoft/amos/blob/HEAD/huggingface/run_glue.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"12cd41e0ee10b80e"}},{"code_sha256_prefix":"8da182308e4e4fd1","entry":"get_optimizer_grouped_parameters","repo":"microsoft/amos","repo_kind":"official","path":"huggingface/run_squad.py","file_url":"https://github.com/microsoft/amos/blob/HEAD/huggingface/run_squad.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"8da182308e4e4fd1"}},{"code_sha256_prefix":"0c9432834249738e","entry":"glue_convert_examples_to_features","repo":"microsoft/amos","repo_kind":"official","path":"huggingface/utils_for_glue.py","file_url":"https://github.com/microsoft/amos/blob/HEAD/huggingface/utils_for_glue.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"0c9432834249738e"}},{"code_sha256_prefix":"4d8fd9c4df4cdbf8","entry":"grad_multiply","repo":"microsoft/amos","repo_kind":"official","path":"fairseq/fairseq/models/amos/model.py","file_url":"https://github.com/microsoft/amos/blob/HEAD/fairseq/fairseq/models/amos/model.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"4d8fd9c4df4cdbf8"}},{"code_sha256_prefix":"3c241ecfe3749a6d","entry":"simple_accuracy","repo":"microsoft/amos","repo_kind":"official","path":"huggingface/utils_for_glue.py","file_url":"https://github.com/microsoft/amos/blob/HEAD/huggingface/utils_for_glue.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"3c241ecfe3749a6d"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}