CORTEXA
← Browse

Tomas Pødenphant Lund

3 papers indexed

openalexZenodo (CERN European Organization for Nuclear Research)2026-07-23

Reading the substrate, not the score: a self-calibrating friction signature that warns of fine-tuning over-fitting and maps when interventions help

Tomas Pødenphant Lund

Early warning of a training-time collapse has so far been read from a model’s internals. This paper shows the same precursor is legible in the model’s own output distribution: per-token entropy variance and lag-1 autocorrelation, computed from the log-probabilities any inference…

View free PDFSource page
openalexZenodo (CERN European Organization for Nuclear Research)2026-07-23

Reading creativity from the inside: a competing-routes view of the diversity–coherence trade-off in language models

Tomas Pødenphant Lund

Raising the temperature does not make a language model more creative, and this paper shows why at the level of the token race: temperature collapses value and coherence without ever raising novelty, while prompts that change which continuations can win raise judged novelty monoto…

View free PDFSource page
openalexOpen MIND2026-07-23

Friction-Guided Inference: Calibrating Correction Strategies and Abstention from Logprob Signals

Tomas Pødenphant Lund

Large language models frequently possess the knowledge needed to answer a question correctly yet commit to the wrong response. This paper presents friction-guided inference, a calibrated inference-time pipeline that uses the model's own logprob distribution — available at zero co…

View free PDFSource page