BOAD shows that coding-agent performance depends less on assembling more agents than on discovering a small team, assigning individual credit, and controlling what each agent needs to remember.
RxnBench reveals why multimodal models that excel on isolated reaction schemes still struggle to read complete chemistry papers reliably.
Why symmetric domain alignment can erase useful information—and how directional simulation offers a safer objective for transfer learning.
A business-focused reading of dynamic data weighting in LLM training, and why selective forgetting may matter more than simply feeding models more tokens.
A multilingual prompt-injection experiment shows why documents must be treated as active attack surfaces—and why apparent resistance in one language may still conceal unstable decisions.
A classroom trial reveals that effective AI tutoring depends less on autonomous intelligence than on diagnostic context, constrained generation, human judgment, and careful measurement.
MIRAGE-VC shows how utility-aware graph retrieval, specialist agents, and adaptive evidence fusion can turn sprawling relationship networks into focused decision-support.
NeuroSPICE recasts circuit simulation as continuous, differentiable function learning—promising easier emerging-device modeling and optimization, but not a faster replacement for SPICE.
I-PERI shows how intervention-driven differences across private datasets can reveal causal directions that ordinary federated learning would discard as inconvenient heterogeneity.
Hindsight Instruction Replay shows how partially compliant model responses can become useful positive training examples without replacing clear binary rewards with ambiguous partial-credit scores.