Mariusz Kulma

End-to-end decision latency of a confidence-gated Laya–phi4 cascade on one consumer GPU

A small, fast text classifier can answer the easy cases itself and pass only the uncertain ones to a large language model. Version 1.0 of this work estimated how much time such a two-stage setup saves; this version measures it while both models run on one consumer graphics card....

Pair and triple correlations of Riemann zeros at low heights: limits of effective-size corrections

To leading order, the Riemann zeros at height T deviate from the random-matrix limit in the same way as the eigenvalues of a random unitary matrix of finite size N_e ∝ log T. For the zeros studied here, this effective matrix is tiny: N_e is about 1 to 5. We ask whether so small a...

Testing six LLM-generated sleep-EEG predictions on open polysomnography: a case study with a pre-frozen protocol and negative results

Language models can write research hypotheses much faster than anyone can test them. Earlier studies have judged such hypotheses by expert ratings, benchmarks with known answers or laboratory experiments. We report a small, fully documented case from a different angle: what happe...

The first reads are not a random sample: archive FASTQ files from aligned submissions can inflate two-million-read telomeric estimates 7- to 10-fold

Whole-genome sequencing files are large, and a common shortcut is to analyse only their first few million reads. That is safe only if the order of reads in the file carries no information. We asked whether the shortcut distorts a rare signal, the share of reads made of telomeric...

Would Proving the Riemann Hypothesis Break RSA?

Would a proof of the Riemann Hypothesis break RSA encryption? The rumour is old. For the known consequences and the algorithms examined here, the answer is no, but the connection is more interesting than the rumour suggests. A hypothesis about zeros of L-functions could touch RSA...

Would Proving the Riemann Hypothesis Break RSA?

Would a proof of the Riemann Hypothesis break RSA encryption? The rumour is old. For the known consequences and the algorithms examined here, the answer is no, but the connection is more interesting than the rumour suggests. A hypothesis about zeros of L-functions could touch RSA...

The first reads are not a random sample: archive FASTQ files from aligned submissions can show head-of-file telomeric enrichment and inflate two-million-read estimates 7- to 10-fold

Whole-genome sequencing files are large, and a common shortcut is to analyse only their first few million reads. That is safe only if the order of reads in the file carries no information. We asked whether the shortcut distorts a rare signal, the share of reads made of telomeric...

The first reads are not a random sample: archive FASTQ files from aligned submissions can show head-of-file telomeric enrichment and inflate two-million-read estimates 7- to 10-fold

Whole-genome sequencing files are large, and a common shortcut is to analyse only their first few million reads. That is safe only if the order of reads in the file carries no information. We asked whether the shortcut distorts a rare signal, the share of reads made of telomeric...

Drift-aware decoding of a repetition code: H7 boundary study, H8 decomposition and H9 co-drifter addendum

Technical notes and complete deposit (code, synthetic data, preregistrations, decision log, independent audits) for three preregistered synthetic studies of drift-aware decoding of a distance-5 repetition code with one drifting data qubit. QP-02-H7, a kill test at equal informati...

Risk-Utility Trade-offs in Selective Decision Gates: A Synthetic Study of Outlier Contamination

A selective decision gate can reduce wrong decisions by withholding actions while losing much of the baseline system's useful output. We examine this trade-off in a classical, four-hypothesis sequential simulator through four archived experimental stages: an exploratory pilot, a...

Risk-Utility Trade-offs in Selective Decision Gates: A Synthetic Study of Outlier Contamination

A selective decision gate can reduce wrong decisions by withholding actions while losing much of the baseline system's useful output. We examine this trade-off in a classical, four-hypothesis sequential simulator through four archived experimental stages: an exploratory pilot, a...