Roberto Infante

Cognitive computing · decision-making in humans and machines

rdji@uw.edu · CV · GitHub · LinkedIn · Twitter

Roberto Infante

I study awareness of unawareness in sequential decision making, with a focus on recognition of insufficiency: how humans and machines recognize that their currently available possibilities may be insufficient before necessarily knowing what is missing. I am interested in what cues from ongoing reasoning support this recognition, how it affects cognitive control, and whether similar mechanisms exist in language models. I build controlled, species-fair tasks that go beyond final accuracy to compare how humans and machines reach an answer.



Clusterwords

Clusterwords is a controlled semantic reasoning task in which humans and machines identify groups of words linked by a common relation (NYT Connections). It lets us manipulate the information and structure of the problem while measuring how possibilities are generated, revised, and committed to over time.

Using Clusterwords, I want to study how information from ongoing reasoning affects recognition of insufficiency, cognitive control, confidence, search, and commitment.

Four words. Pick the two that belong together — then say why.
Play the full game →

Currently engineering an automatic board-generation pipeline for Clusterwords to create novel Connections-like problems at scale and collect reasoning and decision-making data from humans and language models.

Difficulty is manipulated along two independent axes: combinatorial (board size) and semantic (structure, margin, distractors). Boards are machine-generated to reduce overlap with any model's training data.

Working papers

Gentags: Discrete Semantic State for Constraint-Sensitive Decision Pipelines 2025 – 2026

First-author preprint introducing Gentags, a representation that compresses source text into short, evidence-grounded semantic units used as inspectable intermediate state in LLM decision pipelines. In a controlled study isolating representation structure, Gentags raised agreement with full-evidence decisions to 79.5% (vs. 52.3–61.6% for RAKE/YAKE/TF-IDF) and hard-constraint satisfaction to 97.3% (vs. 84.7–89.3%). [PDF]

Other projects

Dynamic Information 2024

An LLM agent that builds and self-updates a knowledge/decision graph with an explicit hypothesis structure for Bayesian-style reasoning. Its limits motivated the shift to studying commitment under uncertainty directly.

Recommendation System & Smart Filters — UW MSIM Capstone (sponsor & lead) 2025 – 2026

Directed two capstone teams applying the Gentags approach to a live product, shipping a recommendation system and categorical/nudging filters now in production.

Reading

A shelf around cognition, computation, and decision-making. Drag the shelf; the laptop on the right holds the papers. Read and the next on the line.

Papers — pick one to open it on the laptop

Background

Previously I built and shipped data-driven products end to end as a technical founder. I am now moving into research in computational cognition.