{"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/neftune-noisy-embeddings-improve-instruction","title":"NEFTune: Noisy Embeddings Improve Instruction Finetuning","arxiv_id":"2310.05914","date":"2023-10-09","proceeding":null,"authors":["Neel Jain","Ping-Yeh Chiang","Yuxin Wen","John Kirchenbauer","Hong-Min Chu","Gowthami Somepalli","Brian R. Bartoldson","Bhavya Kailkhura","Avi Schwarzschild","Aniruddha Saha","Micah Goldblum","Jonas Geiping","Tom Goldstein"],"abstract":"We show that language model finetuning can be improved, sometimes dramatically, with a simple augmentation. NEFTune adds noise to the embedding vectors during training. Standard finetuning of LLaMA-2-7B using Alpaca achieves 29.79% on AlpacaEval, which rises to 64.69% using noisy embeddings. NEFTune also improves over strong baselines on modern instruction datasets. Models trained with Evol-Instruct see a 10% improvement, with ShareGPT an 8% improvement, and with OpenPlatypus an 8% improvement. Even powerful models further refined with RLHF such as LLaMA-2-Chat benefit from additional training with NEFTune.","url_abs":"https://arxiv.org/abs/2310.05914v2","url_pdf":"https://arxiv.org/pdf/2310.05914v2.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":"neftune-noisy-embeddings-improve-instruction","repo_url":"https://github.com/neelsjain/neftune","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"neftune-noisy-embeddings-improve-instruction","repo_url":"https://github.com/akjindal53244/arithmo","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"neftune-noisy-embeddings-improve-instruction","repo_url":"https://github.com/openaccess-ai-collective/axolotl","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"neftune-noisy-embeddings-improve-instruction","repo_url":"https://github.com/rijgersberg/geitje","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"language-modelling","task_name":"Language Modelling"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2310.05914","atlas_url":"https://app.syntology.ai/?focus=2310.05914","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2310.05914"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+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/openaccess-ai-collective/axolotl","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/akjindal53244/arithmo","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/rijgersberg/geitje","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/neelsjain/neftune","reach":null}],"summary":{"ran":3,"ran_violates":1,"ran_draft_wrong":2,"ran_honours":1,"unverified":1},"by_repo_kind":{"official":{"samples":1,"ran":0,"repositories":1},"listed":{"samples":7,"ran":7,"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":"466b972ae1ee4368","entry":"extract_ground_truth_answer","repo":"akjindal53244/arithmo","repo_kind":"listed","path":"eval/gsm8k/gsm8k_compute_metric_zero_shot_CoT.py","file_url":"https://github.com/akjindal53244/arithmo/blob/HEAD/eval/gsm8k/gsm8k_compute_metric_zero_shot_CoT.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"466b972ae1ee4368"}},{"code_sha256_prefix":"d5cb757080457d06","entry":"extract_predcited_answer","repo":"akjindal53244/arithmo","repo_kind":"listed","path":"eval/gsm8k/gsm8k_compute_metric_zero_shot_CoT.py","file_url":"https://github.com/akjindal53244/arithmo/blob/HEAD/eval/gsm8k/gsm8k_compute_metric_zero_shot_CoT.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"d5cb757080457d06"}},{"code_sha256_prefix":"bccdf81fed5534bd","entry":"extract_python_program","repo":"akjindal53244/arithmo","repo_kind":"listed","path":"eval/gsm8k/gsm8k_write_zero_shot_PoT_outputs.py","file_url":"https://github.com/akjindal53244/arithmo/blob/HEAD/eval/gsm8k/gsm8k_write_zero_shot_PoT_outputs.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"bccdf81fed5534bd"}},{"code_sha256_prefix":"20063c675d638a77","entry":"is_equiv","repo":"akjindal53244/arithmo","repo_kind":"listed","path":"eval/MATH/MATH_compute_metric_zero_shot_CoT.py","file_url":"https://github.com/akjindal53244/arithmo/blob/HEAD/eval/MATH/MATH_compute_metric_zero_shot_CoT.py","link_basis":"harvester_set","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"20063c675d638a77"}},{"code_sha256_prefix":"0b14c648516c38a7","entry":"last_boxed_only_string","repo":"akjindal53244/arithmo","repo_kind":"listed","path":"eval/MATH/MATH_compute_metric_zero_shot_CoT.py","file_url":"https://github.com/akjindal53244/arithmo/blob/HEAD/eval/MATH/MATH_compute_metric_zero_shot_CoT.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"0b14c648516c38a7"}},{"code_sha256_prefix":"f3bbe264b05aadd3","entry":"remove_boxed","repo":"akjindal53244/arithmo","repo_kind":"listed","path":"eval/MATH/MATH_compute_metric_zero_shot_CoT.py","file_url":"https://github.com/akjindal53244/arithmo/blob/HEAD/eval/MATH/MATH_compute_metric_zero_shot_CoT.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"f3bbe264b05aadd3"}},{"code_sha256_prefix":"5e4b02e5bc753357","entry":"similar","repo":"akjindal53244/arithmo","repo_kind":"listed","path":"data_prep/prepare_model_traininig_data.py","file_url":"https://github.com/akjindal53244/arithmo/blob/HEAD/data_prep/prepare_model_traininig_data.py","link_basis":"harvester_set","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"5e4b02e5bc753357"}},{"code_sha256_prefix":"903c643febb1d508","entry":"generate_responses_batched","repo":"neelsjain/neftune","repo_kind":"official","path":"experiment_code/eval_generate.py","file_url":"https://github.com/neelsjain/neftune/blob/HEAD/experiment_code/eval_generate.py","link_basis":"first_harvest_node","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":"903c643febb1d508"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}