6 papers
Patch where the text is new: a compression view of where byte-level models should spend compute
Nicolás Vera Zúñiga
Byte-level language models such as the Byte Latent Transformer (BLT) run their large global model once per patch and start a patch where a small model's next-byte entropy is high.…
Unknown is not normal: separating language-model extraction from rule-based decision logic for clinical risk scores
Nicolás Vera Zúñiga
Large language models (LLMs) are increasingly used to compute clinical risk scores from free-text notes. Notes are often incomplete, and treating undocumented findings as normal ca…
Where a Model Sends Its Own Repeated Token
Nicolás Vera Zúñiga
Black-box model identification works by scoring a model's response to natural-language prompts. One line of work feeds models a degenerate input -- their own token, repeated -- to…
What a Cross-Model Fixed-Point Census Can and Cannot Arbitrate About Repetition
Nicolás Vera Zúñiga
Two accounts of neural text degeneration coexist. One locates the cause in the training data -- repetition in the corpus produces repetition in the output, established by training…
Prompt-Model Interaction Reaches the Fixed Points: A deterministic, task-free structural readout -- and the factorizations of it that failed
Nicolás Vera Zúñiga
That a prompt's effect is not a property of the prompt is established: prompts optimised for one model degrade on another, and rankings reorder under neutral reformatting. That evi…
What Iterated Self-Feeding Probes of Language Models Measure, and a test that separates the construction from the model
Nicolás Vera Zúñiga
A growing class of methods probes a language model by feeding it its own output: self-consistency, iterated refinement, agentic loops. We ask what such a probe measures, in a const…