RIKEN

AI agents can now propose atom-level biological hypotheses faster than they can be tested, but trust in an agent-run molecular dynamics simulation rests on a process record, and existing agents leave that record to the language model itself. MDClaw pairs validating tools with a directed acyclic graph (DAG, a graph recording which calculation stages depend on which) that checks stages before execution and records each one; on MDDataBench, which grades against deposited simulations, four open-weight models with MDClaw passed all three runs on 81–100% of 31 test systems. An ablation without the DAG found 11 runs that passed every output check yet had skipped minimization or equilibration, or simulated a different structure.

AI agents can now propose atom-level biological hypotheses faster than they can be tested, but trust in an agent-run molecular dynamics simulation rests on a process record, and existing agents leave that record to the language model itself. MDClaw pairs validating tools with a directed acyclic graph (DAG, a graph recording which calculation stages depend on which) that checks stages before execution and records each one; on MDDataBench, which grades against deposited simulations, four open-weight models with MDClaw passed all three runs on 81–100% of 31 test systems. An ablation without the DAG found 11 runs that passed every output check yet had skipped minimization or equilibration, or simulated a different structure.

Cell tracking has long depended on a person deciding, dataset by dataset, what to annotate, when to train, what went wrong and when to stop. Automahout, an autonomous agent on the Mastodon/ELEPHANT platform from RIKEN, first tries existing detectors; where none fits it annotates reference cells, trains a detector from those sparse annotations, reviews the output against them and feeds corrections back, then links the lineage. On three public datasets it ran with no human input and reached TRA of 0.95–0.96.

Cell tracking has long depended on a person deciding, dataset by dataset, what to annotate, when to train, what went wrong and when to stop. Automahout, an autonomous agent on the Mastodon/ELEPHANT platform from RIKEN, first tries existing detectors; where none fits it annotates reference cells, trains a detector from those sparse annotations, reviews the output against them and feeds corrections back, then links the lineage. On three public datasets it ran with no human input and reached TRA of 0.95–0.96.