Series

Frontier Scout

The upstream of the publication pipeline. An agent reads the global stream of new AI papers, scores each by popularity and code, and verdicts whether the core hypothesis is reproducible on the Spark.

Article №28 observability NIM ~3 hours — 30 min plumbing, ~20 min for the runs themselves, the rest is reading what they show
Frontier Scout

AutoResearchBench on Spark — Two NIMs, One Bench, Two Failure Modes

Two Spark-tuned NIMs run AutoResearchBench's three Deep-Research example questions. Llama-3.1-8B crashes by turn 5-6 on its 8K context; Nemotron-Nano-9B-v2 finishes cleanly at 128K. Both score 0% Accuracy@1 — for completely different reasons.

uses fieldkit.nimfieldkit.evalfieldkit.capabilities

Article №29 inference Foundation ~2 hours — most of it watching vLLM 0.20 build inside an NGC PyTorch container; the runtime+drift diagnosis that follows is the short, sharp half
Frontier Scout

Test-Time Distilling on Spark — Same Compute Envelope, Wider Semantic Reach

ESamp adds a tiny test-time-trained probe to vLLM that converts decoding from lexical resampling into semantic exploration. The runtime is vLLM-native — and that is a Spark catalog-gap story before it is a benchmark.

uses fieldkit.evalfieldkit.capabilities

Article №30 inference Foundation ~2 hours of patching · ~30 minutes of measuring
Frontier Scout

Two Patches Were Six — ESamp Lands at 97.4% on a Patched Spark

Article #2 closed at two patches. Applying them surfaced six — including the silent return-shape adapter that broke the consumer's port. Once cleared, ESamp lands at 97.4% of baseline on patched Qwen 2.5 7B, within 1.4 pp of the paper's reference.

uses fieldkit.evalfieldkit.capabilities

Article №31 inference Foundation ~3 hours of measurement · ~one line of patch
Frontier Scout

Pass@k After the Seventh Patch — Three Shapes ESamp Takes on Spark

Patches were six. The Pass@k harness surfaced a seventh — a one-line slice in the residual tap that only fires once batches shrink mid-run. Once cleared, ESamp takes three shapes: flat on saturated cells, lifting both rates on instruct headroom, and +6.67pp pass@8 on the unsaturated reasoning cell.

uses fieldkit.evalfieldkit.capabilities

Article №32 fine-tuning NeMo ~3 days end-to-end (mostly waiting on rollouts)
Frontier Scout

ClawGym on Spark — A 7B Base, A LoRA Adapter, and the +15 pp the Adapter Earned

ClawGym shipped only a .github profile, so we built the substrate ourselves — persona task synth, sandbox harness, 200-task corpus, LoRA SFT, matched-base eval. The adapter earns +3.8 pp task pass and +15.0 pp per-assertion against its own base. The diagnostic is the lift.

uses fieldkit.nim

Article №33 fine-tuning NeMo ~9 hours wall (34 GRPO steps + two evals)
Frontier Scout

ClawGym GRPO on Spark — Closing the Loop the SFT Adapter Couldn't

Phase 5 SFT taught the agent to keep working but never to stop. 34 GRPO steps with a shaped reward unlearn the failure mode — same model, same base, same LoRA-init, but task_complete climbs 0/158 → 154/158, mean turns drop 12 → 5, and per-assertion still inches up +3.1 pp.
Article №34 fine-tuning NeMo ~18.5 hours wall (50 T²PO steps + three evals)
Frontier Scout

T²PO on Spark — When the Training Pool Says 28/32 and Held-out Says 9/158

T²PO's two deltas on the Phase 6 ClawGym harness: mean turns 5.00 → 4.61, task_complete 154/158, but the per-assertion ceiling stays flat at 47.7%. The strongest training-side step (45) is the worst held-out checkpoint — pool saturation lies on a single Spark.

uses fieldkit.capabilitiesfieldkit.evalfieldkit.training