Modeling Reference Production as the Probabilistic Combination of Multiple Perspectives
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
While speakers have been shown to adapt to the knowledge state of their addressee in choosing referring expressions, they often also show some egocentric tendencies. The current paper aims to provide an explanation for this "mixed" behavior by presenting a model that derives such patterns from the probabilistic combination of both the speaker's and the addressee's perspectives. To test our model, we conducted a language production experiment, in which participants had to refer to objects in a context that also included a visually misleading object (e.g., a crayon shaped like a Lego brick) whose function was either known to both partners, or known just to the speaker but not the addressee. Modeling results indicate that the experimental findings cannot be explained by assuming that speakers tailor a referring expression solely to their own perspective or to the perspective of their addressee. Instead, accounting for the behavioral pattern requires an approach where both perspectives influence the choice of referring expressions. Nevertheless, in our situation, speakers consider their partner's perspective more than their own.
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 | 0.001 | 0.004 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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