Search papers, labs, and topics across Lattice.
The CoTu team tackled the EXACT 2026 competition by developing a neuro-symbolic Program-of-Thought pipeline that integrates logical reasoning and multi-step physics problem solving using a 4B parameter model. Their approach generates explainable answers by encoding regulations in Z3 and producing numerical Python for physics, all while adhering to a strict 60-second query limit through advanced routing and fine-tuning techniques. This resulted in a perfect score on the physics task and the highest technical score overall, demonstrating that smaller models can achieve high reasoning accuracy when grounded in symbolic solvers.
Achieving a perfect score in physics with a 4B model proves that explainability and reasoning power can coexist even in smaller architectures.
Transparent educational question answering asks for answers that are not only correct but explainable, and doing so with small models rules out the reasoning power of the largest proprietary systems. The EXACT 2026 competition poses this problem concretely: open-weight language models of at most 8B parameters, self-hosted, with a natural-language explanation for every answer. It pairs two tasks: logical reasoning over university regulations, and multi-step physics problem solving. We describe the system that team \cotu{} developed to address both, a neuro-symbolic Program-of-Thought pipeline in which a 4B backbone writes a program rather than stating an answer directly: for regulation queries it emits a Z3 encoding whose entailment verdict grounds the deduction, and for physics it emits numerical Python, both wrapped in a shared self-correction loop and a unified explained-JSON output. Answer-type routing, distillation-based task fine-tuning, and a latency-aware serving stack -- SGLang with speculative decoding -- keep the system within the 60-second per-query limit. The system achieved a \textbf{perfect score} on the physics task in both automated selection rounds and obtained the \textbf{highest final-round technical score} of any team -- $13.44/15$, combining automated answer evaluation with expert-judged reasoning depth -- with the equally weighted presentation score included, \cotu{} placed 3rd overall. Grounding answers in a symbolic solver yields correct, verifiable deductions at the 4B scale, and the residual difficulty lies in premise selection rather than the deduction itself.
