{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,30]],"date-time":"2026-09-30T06:56:20Z","timestamp":1790751380593,"version":"4.1.0"},"reference-count":35,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,9,30]],"date-time":"2026-09-30T00:00:00Z","timestamp":1790726400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Neurosci."],"abstract":"<jats:sec>\n                    <jats:title>Introduction<\/jats:title>\n                    <jats:p>\n                      The rapid proliferation of large language models (LLMs) has intensified the need for robust evaluation frameworks to align artificial intelligence (AI) behavior with human expectations. While current reinforcement learning from human feedback (RLHF) strategies heavily rely on explicit,\n                      <jats:italic>post-hoc<\/jats:italic>\n                      self-reports, these metrics are fundamentally limited by subjective biases and user non-compliance.\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>To address this, the current study presents a novel brain-computer interface (BCI) framework that systematically characterizes the continuous, implicit neural correlates of user satisfaction and dissatisfaction during a reading-based evaluative task involving pre-validated LLM-generated question-answer pairs. Sixty-four-channel electroencephalography (EEG) data were recorded from 30 participants while they read and evaluated various LLM-generated responses. To capture genuine subjective experiences, individual trials were relabeled based on post-trial 5-point Likert scale ratings into five distinct evaluative states, and spatiotemporal spectral analyses were performed. A ShallowFBCSPNet architecture was further used to evaluate single-trial classification between the very satisfied and very dissatisfied conditions under a subject-mixed cross-validation scheme.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Spatiotemporal spectral analyses revealed distinct neural oscillatory dynamics that temporally dissociate satisfaction and dissatisfaction. When reading highly satisfactory responses, a pronounced, sustained enhancement of low-frequency delta power was observed within the bilateral frontotemporal areas during the middle (4\u20138 s) and late (8\u201312 s) reading intervals, indexing high-order cognitive consolidation and reward processing. Conversely, processing highly dissatisfactory responses led to a localized suppression of high-frequency gamma power in the right frontotemporal region during the early interval (0\u20134 s), reflecting an immediate affective appraisal and expectation violation. Utilizing the ShallowFBCSPNet architecture, the system successfully discriminated between the \u201cvery satisfied\u201d and \u201cvery dissatisfied\u201d states, achieving an average accuracy of 81% under a subject-mixed cross-validation scheme.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Discussion<\/jats:title>\n                    <jats:p>These findings demonstrate that user evaluative states are encoded via distinct frequency-specific cortical networks over differing temporal windows. By bridging cognitive neuroscience and artificial intelligence, this study establishes the conceptual and empirical foundation for passive BCI-driven implicit feedback channels, laying groundwork toward future real-time paradigms for next-generation LLM alignment; the classification result should be interpreted as a feasibility demonstration rather than an estimate of real-world decoding accuracy.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fnins.2026.1934150","type":"journal-article","created":{"date-parts":[[2026,9,30]],"date-time":"2026-09-30T06:06:31Z","timestamp":1790748391000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Decoding user satisfaction during reading of pre-defined question\u2013answer pairs: BCI approach using delta and gamma EEG"],"prefix":"10.3389","volume":"20","author":[{"given":"Wanjoo","family":"Park","sequence":"first","affiliation":[{"name":"Engineering Division, New York University Abu Dhabi","place":["Abu Dhabi, United Arab Emirates"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Salma","family":"Mansour","sequence":"additional","affiliation":[{"name":"Engineering Division, New York University Abu Dhabi","place":["Abu Dhabi, United Arab Emirates"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Laehyun","family":"Kim","sequence":"additional","affiliation":[{"name":"Bionics Research Center, Korea Institute of Science and Technology (KIST)","place":["Seoul, Republic of Korea"]},{"name":"Department of HY-KIST Bio-convergence, Hanyang University","place":["Seoul, Republic of Korea"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mohamad","family":"Eid","sequence":"additional","affiliation":[{"name":"Engineering Division, New York University Abu Dhabi","place":["Abu Dhabi, United Arab Emirates"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,9,30]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"1655","DOI":"10.3389\/fpsyg.2018.01655","article-title":"Beyond self-report: a review of physiological and neuroscientific methods to investigate consumer behavior","volume":"9","author":"Bell","year":"2018","journal-title":"Front. Psychol"},{"key":"B2","doi-asserted-by":"publisher","first-page":"626","DOI":"10.1111\/psyp.12390","article-title":"Time-frequency theta and delta measures index separable components of feedback processing in a gambling task","volume":"52","author":"Bernat","year":"2015","journal-title":"Psychophysiology"},{"key":"B3","doi-asserted-by":"publisher","first-page":"1114","DOI":"10.1109\/TBME.2019.2930186","article-title":"Evaluation of artifact subspace reconstruction for automatic artifact components removal in multi-channel EEG recordings","volume":"67","author":"Chang","year":"2019","journal-title":"IEEE Trans. Biomed. Eng"},{"key":"B4","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3743127","article-title":"Rlhf deciphered: a critical analysis of reinforcement learning from human feedback for llms","volume":"58","author":"Chaudhari","year":"2025","journal-title":"ACM Comput. Surv"},{"key":"B5","doi-asserted-by":"publisher","first-page":"107848","DOI":"10.1016\/j.neunet.2025.107848","article-title":"EEG emotion copilot: Optimizing lightweight llms for emotional EEG interpretation with assisted medical record generation","volume":"192","author":"Chen","year":"2025","journal-title":"Neural Netw"},{"key":"B6","doi-asserted-by":"publisher","first-page":"5","DOI":"10.1016\/S0987-7053(00)00055-1","article-title":"Removal of ocular artifact from the EEG: a review","volume":"30","author":"Croft","year":"2000","journal-title":"Clin. Neurophysiol"},{"key":"B7","first-page":"50750","article-title":"\u201cSafe rlhf: safe reinforcement learning from human feedback,\u201d","author":"Dai","year":"2024","journal-title":"International conference on learning representations"},{"key":"B8","doi-asserted-by":"publisher","first-page":"2372","DOI":"10.1038\/s41598-023-27528-0","article-title":"EEG is better left alone","volume":"13","author":"Delorme","year":"2023","journal-title":"Sci. Rep"},{"key":"B9","doi-asserted-by":"publisher","first-page":"9","DOI":"10.1016\/j.jneumeth.2003.10.009","article-title":"EEGlab: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis","volume":"134","author":"Delorme","year":"2004","journal-title":"J. Neurosci. Methods"},{"key":"B10","doi-asserted-by":"publisher","first-page":"995","DOI":"10.1093\/scan\/nsaa131","article-title":"EEG response to game-craving according to personal preference for games","volume":"16","author":"Ha","year":"2021","journal-title":"Soc. Cogn. Affect. Neurosci"},{"key":"B11","doi-asserted-by":"publisher","first-page":"347","DOI":"10.1016\/j.tics.2004.06.006","article-title":"Cognitive functions of gamma-band activity: memory match and utilization","volume":"8","author":"Herrmann","year":"2004","journal-title":"Trends Cogn. Sci"},{"key":"B12","doi-asserted-by":"publisher","first-page":"338","DOI":"10.3389\/fnhum.2013.00338","article-title":"Dissociating neuronal gamma-band activity from cranial and ocular muscle activity in EEG","volume":"7","author":"Hipp","year":"2013","journal-title":"Front. Hum. Neurosci"},{"key":"B13","doi-asserted-by":"publisher","first-page":"401","DOI":"10.1111\/j.1467-9280.1993.tb00589.x","article-title":"When more pain is preferred to less: adding a better end","volume":"4","author":"Kahneman","year":"1993","journal-title":"Psychol. Sci"},{"key":"B14","doi-asserted-by":"publisher","first-page":"357","DOI":"10.1109\/TOH.2018.2804952","article-title":"A vibrotactile alarm system for pleasant awakening","volume":"11","author":"Korres","year":"2018","journal-title":"IEEE Trans. Haptics"},{"key":"B15","doi-asserted-by":"publisher","first-page":"3596","DOI":"10.1109\/TAFFC.2025.3605833","article-title":"Libeer: a comprehensive benchmark and algorithm library for EEG-based emotion recognition","volume":"16","author":"Liu","year":"2025","journal-title":"IEEE Trans. Affect. Comput"},{"key":"B16","doi-asserted-by":"publisher","first-page":"zsad241","DOI":"10.1093\/sleep\/zsad241","article-title":"Artifact subspace reconstruction: a candidate for a dream solution for EEG studies, sleep or awake","volume":"46","author":"Miyakoshi","year":"2023","journal-title":"Sleep"},{"key":"B17","doi-asserted-by":"publisher","first-page":"168","DOI":"10.1145\/3704137.3704182","article-title":"\u201cMetrics of success: evaluating user satisfaction in ai chatbots,\u201d","author":"M\u00f8ller","year":"2024","journal-title":"Proceedings of the 2024 8th international conference on advances in artificial intelligence"},{"key":"B18","doi-asserted-by":"publisher","first-page":"138","DOI":"10.3389\/fnhum.2013.00138","article-title":"High-frequency brain activity and muscle artifacts in meg\/EEG: a review and recommendations","volume":"7","author":"Muthukumaraswamy","year":"2013","journal-title":"Front. Hum. Neurosci"},{"key":"B19","doi-asserted-by":"publisher","first-page":"282","DOI":"10.1002\/pra2.456","article-title":"\u201cNeural correlates of realisation of satisfaction in a successful search process,\u201d","author":"Paisalnan","year":"2021","journal-title":"Proceedings of the association for information science and technology"},{"key":"B20","doi-asserted-by":"publisher","first-page":"443","DOI":"10.1007\/978-3-031-25891-6_34","article-title":"\u201cNeural correlates of satisfaction of an information need,\u201d","author":"Paisalnan","year":"2022","journal-title":"International conference on machine learning, optimization, and data science"},{"key":"B21","doi-asserted-by":"publisher","first-page":"14267","DOI":"10.1038\/s41598-024-65289-6","article-title":"EEG correlates to perceived urgency elicited by vibration stimulation of the upper body","volume":"14","author":"Park","year":"2024","journal-title":"Sci. Rep"},{"key":"B22","doi-asserted-by":"publisher","first-page":"22324","DOI":"10.1109\/ACCESS.2018.2827023","article-title":"Gamma EEG correlates of haptic preferences for a dial interface","volume":"6","author":"Park","year":"2018","journal-title":"IEEE Access"},{"key":"B23","doi-asserted-by":"publisher","first-page":"682113","DOI":"10.3389\/fnins.2021.682113","article-title":"Neural coding of vibration intensity","volume":"15","author":"Park","year":"2021","journal-title":"Front. Neurosci"},{"key":"B24","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3772318.3791835","article-title":"\u201cEmulating aggregate human choice behavior and biases with gpt conversational agents,\u201d","author":"Pilli","year":"2026","journal-title":"Proceedings of the 2026 CHI conference on human factors in computing systems"},{"key":"B25","doi-asserted-by":"publisher","first-page":"549","DOI":"10.1007\/978-3-031-25891-6_41","article-title":"\u201cRevisiting neurological aspects of relevance: an EEG study,\u201d","author":"Pinkosova","year":"2022","journal-title":"International conference on machine learning, optimization, and data science"},{"key":"B26","doi-asserted-by":"publisher","first-page":"100295","DOI":"10.1016\/j.chbr.2023.100295","article-title":"Moderating effects of self-perceived knowledge in a relevance assessment task: an EEG study","volume":"11","author":"Pinkosova","year":"2023","journal-title":"Comput. Hum. Behav. Rep"},{"key":"B27","doi-asserted-by":"publisher","first-page":"879","DOI":"10.1037\/0021-9010.88.5.879","article-title":"Common method biases in behavioral research: a critical review of the literature and recommended remedies","volume":"88","author":"Podsakoff","year":"2003","journal-title":"J. Appl. Psychol"},{"key":"B28","doi-asserted-by":"publisher","first-page":"031507","DOI":"10.1063\/5.0047237","article-title":"Mind the gap: state-of-the-art technologies and applications for EEG-based brain-computer interfaces","volume":"5","author":"Portillo-Lara","year":"2021","journal-title":"APL Bioeng"},{"key":"B29","doi-asserted-by":"publisher","first-page":"1658756","DOI":"10.3389\/fpsyg.2025.1658756","article-title":"Perceptual load modulates the delta oscillation and the contribution of delta oscillation to reward positivity during feedback valence encoding","volume":"16","author":"Qi","year":"2025","journal-title":"Front. Psychol"},{"key":"B30","doi-asserted-by":"publisher","first-page":"5391","DOI":"10.1002\/hbm.23730","article-title":"Deep learning with convolutional neural networks for EEG decoding and visualization","volume":"38","author":"Schirrmeister","year":"2017","journal-title":"Hum. Brain Mapp"},{"key":"B31","doi-asserted-by":"publisher","first-page":"1623331","DOI":"10.3389\/fnhum.2025.1623331","article-title":"High gamma EEG responses to emotional stimuli in virtual reality: insights from local activation and distributed characteristics","volume":"19","author":"Xiao","year":"2025","journal-title":"Front. Hum. Neurosci"},{"key":"B32","doi-asserted-by":"publisher","first-page":"114289","DOI":"10.52202\/079017-3630","article-title":"\u201cCal-dpo: calibrated direct preference optimization for language model alignment,\u201d","author":"Xiao","year":"2024","journal-title":"Advances in neural information processing systems"},{"key":"B33","doi-asserted-by":"publisher","first-page":"2012627","DOI":"10.1007\/978-981-96-6259-3","article-title":"A survey of large language models","volume":"20","author":"Zhao","year":"2026","journal-title":"Front. Comput. Sci"},{"key":"B34","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3555597","article-title":"\u201cBias-aware design for informed decisions: Raising awareness of self-selection bias in user ratings and reviews,\u201d","author":"Zhu","year":"2022","journal-title":"Proceedings of the ACM on human-computer interaction"},{"key":"B35","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/BCI69045.2026.11435049","article-title":"\u201cDecoding workload and agreement from EEG during spoken dialogue with conversational AI,\u201d","author":"Zidar","year":"2026","journal-title":"2026 14th international conference on brain-computer interface (BCI)"}],"container-title":["Frontiers in Neuroscience"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnins.2026.1934150\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,30]],"date-time":"2026-09-30T06:06:32Z","timestamp":1790748392000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnins.2026.1934150\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,30]]},"references-count":35,"alternative-id":["10.3389\/fnins.2026.1934150"],"URL":"https:\/\/doi.org\/10.3389\/fnins.2026.1934150","relation":{},"ISSN":["1662-453X"],"issn-type":[{"value":"1662-453X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,30]]},"article-number":"1934150"}}