Decoding and expressing brain activity in a comprehensible form is a challenging frontier in AI. This paper presents Thought2Text, which uses instruction-tuned Large Language Models (LLMs) fine-tuned with EEG data to achieve this goal. The approach involves three stages: (1) training an EEG encoder for visual feature extraction, (2) fine-tuning LLMs on image and text data, enabling multimodal description generation, and (3) further fine-tuning on EEG embeddings to generate text directly from EEG during inference. Experiments on a public EEG dataset collected for six subjects with image stimuli and text captions demonstrate the efficacy of multimodal LLMs (LLaMA-v3, Mistral-v0.3, Qwen2.5), validated using traditional language generation evaluation metrics, as well as fluency and adequacy measures. This approach marks a significant advancement towards portable, low-cost”thoughts-to-text”technology with potential applications in both neuroscience and natural language processing.<\/p>\n","protected":false},"excerpt":{"rendered":"
Decoding and expressing brain activity in a comprehensible form is a challenging frontier in AI. This paper presents Thought2Text, which uses instruction-tuned Large Language Models (LLMs) fine-tuned with EEG data to achieve this goal. The approach involves three stages: (1) training an EEG encoder for visual feature extraction, (2) fine-tuning LLMs on image and text […]<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr-author-ordering":[{"type":"text","value":"Abhijit Mishra","user_id":0},{"type":"text","value":"Shreya Shukla","user_id":0},{"type":"user_nicename","value":"Jose Garrido Torres","user_id":"43904"},{"type":"text","value":"Jacek Gwizdka","user_id":0},{"type":"text","value":"Shounak Roychowdhury","user_id":0}],"msr_publishername":"","msr_publisher_other":"","msr_booktitle":"","msr_chapter":"","msr_edition":"","msr_editors":"","msr_how_published":"arXiv","msr_isbn":"","msr_issue":"","msr_journal":"","msr_number":"","msr_organization":"","msr_pages_string":"","msr_page_range_start":"","msr_page_range_end":"","msr_series":"","msr_volume":"","msr_copyright":"","msr_conference_name":"","msr_doi":"10.18653\/v1\/2025.findings-naacl.207","msr_arxiv_id":"2410.07507","msr_mag_id":"","msr_other_authors":"","msr_other_contributors":"","msr_speaker":"","msr_award":"","msr_affiliation":"","msr_institution":"","msr_host":"","msr_version":"","msr_duration":"","msr_release_tracker_id":"","msr_highlight_type":"","msr_date_display_format":"","msr_main_download_label":"","msr_external_link_label":"","msr_doi_label":"","msr_published_date":"2024-10-10","msr_startdate":"","msr_presentation_date":"","msr_highlight_text":"","msr_notes":"","msr_longbiography":"","msr_publicationurl":"","msr_external_url":"","msr_secondary_video_url":"","msr_conference_url":"","msr_journal_url":"","msr_year":2024,"msr_month":10,"msr_day":10,"msr_microsoftintellectualproperty":false,"msr_pub_id":"c38c3d6e148dc35f22dc6e9b1b7b224f8fe6efca","msr_publication_uploader":[{"type":"url","title":"https:\/\/dblp.org\/rec\/journals\/corr\/abs-2410-07507.html","label_id":243109,"id":false,"viewUrl":false},{"type":"url","title":"https:\/\/arxiv.org\/abs\/2410.07507","label_id":243109,"id":false,"viewUrl":false},{"type":"url","title":"https:\/\/aclanthology.org\/2025.findings-naacl.207.pdf","label_id":243109,"id":false,"viewUrl":false}],"msr_related_uploader":[],"msr_original_fields_of_study":[],"msr_s2_paper_id":"c38c3d6e148dc35f22dc6e9b1b7b224f8fe6efca","msr_s2_pdf_url":"https:\/\/aclanthology.org\/2025.findings-naacl.207.pdf","msr_citation_count_updated":"","msr_citation_count":36,"msr_influential_citations":4,"msr_reference_count":43,"msr_s2_open_access":true,"msr_s2_author_ids":[],"msr_pub_ids":[{"provider":"s2","id":"c38c3d6e148dc35f22dc6e9b1b7b224f8fe6efca"},{"provider":"doi","id":"10.18653\/v1\/2025.findings-naacl.207"},{"provider":"arxiv","id":"2410.07507"},{"provider":"dblp","id":"journals\/corr\/abs-2410-07507"},{"provider":"corpusid","id":"273233909"}],"msr_hide_image_in_river":0,"footnotes":""},"msr-research-highlight":[],"research-area":[13556],"msr-publication-type":[270373],"msr-publisher":[],"msr-publication-cta":[],"msr-focus-area":[],"msr-locale":[268875],"msr-post-option":[],"msr-field-of-study":[246691],"msr-conference":[270414],"msr-journal":[],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-1178914","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-area-artificial-intelligence","msr-locale-en_us","msr-field-of-study-computer-science"],"msr_publishername":"","msr_edition":"","msr_affiliation":"","msr_published_date":"2024-10-10","msr_host":"","msr_duration":"","msr_version":"","msr_speaker":"","msr_other_contributors":"","msr_booktitle":"","msr_pages_string":"","msr_chapter":"","msr_isbn":"","msr_journal":"","msr_volume":"","msr_number":"","msr_editors":"","msr_series":"","msr_issue":"","msr_organization":"","msr_how_published":"arXiv","msr_notes":"","msr_highlight_text":"","msr_release_tracker_id":"","msr_original_fields_of_study":"","msr_download_urls":"","msr_external_url":"","msr_secondary_video_url":"","msr_longbiography":"","msr_microsoftintellectualproperty":0,"msr_main_download":"","msr_publicationurl":"","msr_doi":"10.18653\/v1\/2025.findings-naacl.207","msr_publication_uploader":[{"type":"url","title":"https:\/\/dblp.org\/rec\/journals\/corr\/abs-2410-07507.html","label_id":243109,"id":false,"viewUrl":false},{"type":"url","title":"https:\/\/arxiv.org\/abs\/2410.07507","label_id":243109,"id":false,"viewUrl":false},{"type":"url","title":"https:\/\/aclanthology.org\/2025.findings-naacl.207.pdf","label_id":243109,"id":false,"viewUrl":false}],"msr_related_uploader":[],"msr_citation_count":36,"msr_citation_count_updated":"","msr_s2_paper_id":"c38c3d6e148dc35f22dc6e9b1b7b224f8fe6efca","msr_influential_citations":4,"msr_reference_count":43,"msr_arxiv_id":"2410.07507","msr_s2_author_ids":[],"msr_s2_open_access":true,"msr_s2_pdf_url":"https:\/\/aclanthology.org\/2025.findings-naacl.207.pdf","msr_attachments":[],"msr-author-ordering":[{"type":"text","value":"Abhijit Mishra","user_id":0,"rest_url":false},{"type":"text","value":"Shreya Shukla","user_id":0,"rest_url":false},{"type":"user_nicename","value":"Jose Garrido Torres","user_id":43904,"rest_url":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=Jose Garrido Torres"},{"type":"text","value":"Jacek Gwizdka","user_id":0,"rest_url":false},{"type":"text","value":"Shounak Roychowdhury","user_id":0,"rest_url":false}],"msr_impact_theme":[],"msr_research_lab":[],"msr_event":[],"msr_group":[],"msr_project":[],"publication":[],"video":[],"msr-tool":[],"msr_publication_type":"misc","related_content":[],"_links":{"self":[{"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/1178914","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item"}],"about":[{"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-research-item"}],"version-history":[{"count":3,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/1178914\/revisions"}],"predecessor-version":[{"id":1179114,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/1178914\/revisions\/1179114"}],"wp:attachment":[{"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/media?parent=1178914"}],"wp:term":[{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=1178914"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=1178914"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=1178914"},{"taxonomy":"msr-publisher","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-publisher?post=1178914"},{"taxonomy":"msr-publication-cta","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-cta?post=1178914"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=1178914"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=1178914"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=1178914"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=1178914"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=1178914"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=1178914"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=1178914"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/find.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=1178914"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}