Generalization versus contextualization in automatic evaluation.
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Research has shown that automatic evaluations can be highly robust and difficult to change, highly malleable and easy to change, and highly context dependent. We tested a representational account of these disparate findings, which specifies the conditions under which automatic evaluations reflect (a) initially acquired information, (b) subsequently acquired, counterattitudinal information, or (c) a mixture of both. The account postulates that attention to contextual cues during the encoding of evaluative information determines whether this information is stored in a context-free representation or a contextualized representation. To the extent that attention to context cues is low during the encoding of initial information but is enhanced by exposure to expectancy-violating counterattitudinal information, initial experiences are stored in context-free representations, whereas counterattitudinal experiences are stored in contextualized representations. Hence, automatic evaluations tend to reflect the valence of counterattitudinal information only in the context in which this information was learned (occasion setting) and the valence of initial experiences in any other context (renewal effect). Four experiments confirmed these predictions, additionally showing that (a) the impact of initial experiences was reduced for automatic evaluations in novel contexts when context salience during the encoding of initial information was enhanced, (b) context effects were eliminated altogether when context salience during the encoding of counterattitudinal information was reduced, and (c) enhanced context salience during the encoding of counterattitudinal information produced context-dependent automatic evaluations even when there was no contingency between valence and contextual cues. Implications for automatic evaluation, learning theory, and interventions in applied settings are discussed.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.008 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it