TRACES

“Helping humanity discover the unknown is the ultimate value of AI to humankind.”

Mr. Tianqiao Chen, Founder & CEO of Apodex

TRACES is a new paradigm for Discoverative AI.
It is built not to measure how well AI answers known questions, but how far it can push the frontier of the unknown.

TRACES is a new paradigm for Discoverative AI.
It is built not to measure how well AI answers known questions, but how far it can push the frontier of the unknown.

AAV Capsid
Four tasks · 18 episodes · 11 solvers evaluated
Drug Repurposing
Public-100 · three-seed mean ± sample SD
Clinical Trials
Gate × rigor across three ambiguity episodes
LLM Engineering
All eleven boards, cell for cell
Live Benchmark
1,182 trajectories · 17 environments · outcome and HDS6
AAV Capsid
Four tasks · 18 episodes · 11 solvers evaluated
Drug Repurposing
Public-100 · three-seed mean ± sample SD
Clinical Trials
Gate × rigor across three ambiguity episodes
LLM Engineering
All eleven boards, cell for cell
Live Benchmark
1,182 trajectories · 17 environments · outcome and HDS6
AAV Capsid
Four tasks · 18 episodes · 11 solvers evaluated
Drug Repurposing
Public-100 · three-seed mean ± sample SD
Clinical Trials
Gate × rigor across three ambiguity episodes
LLM Engineering
All eleven boards, cell for cell
Live Benchmark
1,182 trajectories · 17 environments · outcome and HDS6
AAV Capsid
Four tasks · 18 episodes · 11 solvers evaluated
Drug Repurposing
Public-100 · three-seed mean ± sample SD
Clinical Trials
Gate × rigor across three ambiguity episodes
LLM Engineering
All eleven boards, cell for cell
Live Benchmark
1,182 trajectories · 17 environments · outcome and HDS6
AAV Capsid
Four tasks · 18 episodes · 11 solvers evaluated
Drug Repurposing
Public-100 · three-seed mean ± sample SD
Clinical Trials
Gate × rigor across three ambiguity episodes
LLM Engineering
All eleven boards, cell for cell
Live Benchmark
1,182 trajectories · 17 environments · outcome and HDS6
AAV Capsid
Four tasks · 18 episodes · 11 solvers evaluated
Drug Repurposing
Public-100 · three-seed mean ± sample SD
Clinical Trials
Gate × rigor across three ambiguity episodes
LLM Engineering
All eleven boards, cell for cell
Live Benchmark
1,182 trajectories · 17 environments · outcome and HDS6
AAV Capsid
Four tasks · 18 episodes · 11 solvers evaluated
Drug Repurposing
Public-100 · three-seed mean ± sample SD
Clinical Trials
Gate × rigor across three ambiguity episodes
LLM Engineering
All eleven boards, cell for cell
Live Benchmark
1,182 trajectories · 17 environments · outcome and HDS6
AAV Capsid
Four tasks · 18 episodes · 11 solvers evaluated
Drug Repurposing
Public-100 · three-seed mean ± sample SD
Clinical Trials
Gate × rigor across three ambiguity episodes
LLM Engineering
All eleven boards, cell for cell
Live Benchmark
1,182 trajectories · 17 environments · outcome and HDS6
AAV Capsid
Four tasks · 18 episodes · 11 solvers evaluated
Drug Repurposing
Public-100 · three-seed mean ± sample SD
Clinical Trials
Gate × rigor across three ambiguity episodes
LLM Engineering
All eleven boards, cell for cell
Live Benchmark
1,182 trajectories · 17 environments · outcome and HDS6