Research Engineer
Atul Saxena
I build the environments models are trained in, and the tools that show what they learned.
- RL Systems
- Interpretability
- ML Infrastructure
Research Engineer
Atul Saxena
I build the environments models are trained in, and the tools that show what they learned.
Atul Saxena, Pushp Kharat
In plain terms
Instead of a model producing an answer you have to trust, LEMMA produces the derivation: every rewrite is a named rule, every intermediate state is checked, and the verification status is explicit — checked, heuristic, unverified, or unsupported are four different outcomes, not one.
LEMMA is a neuro-symbolic system for long-horizon symbolic mathematical rewriting. It combines a rule engine, guardrail-filtered action generation, evidence-typed verification and learned policy guidance to transform a starting expression into a target through individually validated steps. Six search and decoding methods are evaluated on a frozen 120-problem benchmark.