Lingering Neural Representations of Past Task Features Adversely Affect Future Behavior
DOI: 10.1523/jneurosci.0464-22.2022
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This study investigates the neural mechanisms underlying the N-2 repetition cost (N2RC), a behavioral phenomenon where reaction times are slower when a task repeats after an intervening switch, provided the response requirement has changed. The research addresses a debate regarding whether this cost stems from the retrieval of outdated "event files"—conjunctive neural representations linking context, stimuli, and actions—or from the inhibition of the previous task set. The authors hypothesize that stronger lingering neural representations of past task features adversely affect future behavior when action requirements change. To test this, 35 healthy adults performed a task-switching paradigm while undergoing electroencephalography (EEG) recordings. Participants responded to visual targets based on cue shapes that dictated specific response mappings. The experimental design compared trial triplets where the task repeated after a switch (ABA) with those where it did not (CBA). The critical comparison involved ABA trials where the response requirement switched between the first and third trials (ABA-S) versus CBA trials. The researchers employed multivariate pattern analysis and representational similarity analysis (RSA) on EEG data to decode the strength of conjunctive representations (event files) for each trial. These neural decoding metrics were then linked to behavioral performance using linear mixed-effects modeling. The results replicated previous behavioral findings, showing significant reaction time costs and reduced accuracy for ABA-S trials compared to CBA trials. Crucially, the neural analysis revealed that the magnitude of this behavioral cost was directly proportional to the strength of the conjunctive representation formed during the preceding exposure to the same task (trial N-2). Specifically, stronger neural conjunctions on trial N-2 predicted slower reaction times on trial N in ABA-S conditions. This effect was specific to the conjunctive representation; individual task features (cue, target, or response alone) did not predict the cost. Furthermore, the neural representation from the immediately preceding trial (N-1) did not influence trial N performance. Exploratory cross-correlation analyses suggested that strong past event files facilitate the formation of new conjunctions when task sets match but impair formation when they mismatch, with the strength of this interference correlating with individual N2RC magnitudes. These findings provide direct neurophysiological evidence that the automatic formation of conjunctive event files can have detrimental effects on future behavior when task rules change. The study confirms that lingering neural representations of past task features are a primary driver of the N-2 repetition cost, rather than mere task-set inhibition. This demonstrates that neural decoding of complex task-set representations can predict behavioral outcomes beyond the current trial, offering new insights into how the brain organizes and retrieves episodic information during complex cognition, planning, and problem-solving.
Key finding
The strength of conjunctive neural representations formed during a previous task exposure directly predicts increased reaction time costs on subsequent trials when the required response changes.
Methodology
lab_experiment
Sample size: 35
Provenance
The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via author_sweep_intake on 2026-05-28.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | author_sweep | — | — | 2 | 2026-05-28 |
| archive | success | canonical_url | — | — | 1 | 2026-06-04 |
| extract | success | cached | — | — | 3 | 2026-06-10 |
| clean | success | clean | — | — | 1 | 2026-06-04 |
| chunk | success | chunk | — | — | 1 | 2026-06-04 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-04 |
| enrich | success | — | — | — | 1 | 2026-05-28 |
| promote | success | — | — | — | 1 | 2026-06-04 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 2 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 15 | 2026-06-11 |
| verify | success | — | — | — | 2 | 2026-06-10 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.