Tag: Inference
Glitches can mildly bias EMRI inference in LISA
What the study found Streams of transient, non-Gaussian noise artifacts called glitches can bias inference for extreme mass ratio inspirals (EMRIs) in the Laser Interferometer Space Antenna (LISA), but the effect is relatively small when the glitches are moderately mitigated. The study found that EMRI inference is more robust to glitches than inference for some…
Higher-order statistics improve constraints on reionization history
—
in AstrophysicsWhat the study found Betti numbers, a non-Gaussian statistic that captures topological features, were on average more informative than the power spectra for constraining the average neutral hydrogen fraction. The bispectrum was described as providing limited constraints, but combining higher-order statistics with the cylindrical power spectrum improved the mean figure of merit. Why the authors…
Continuous-time sampler handles unknown-dimensional Bayesian models
—
in MathematicsWhat the study found The paper presents samsara, a continuous-time Markov chain Monte Carlo (CTMCMC) sampler designed for Bayesian inference when the number of parameters is unknown. The authors report that it achieves automatic acceptance of trans-dimensional moves and high sampling efficiency. Why the authors say this matters The authors say this matters because many…
Survey maps graph roles in retrieval-augmented generation
Survey of graph-based techniques in retrieval-augmented generation systems, examining their roles in database construction, algorithms, and reasoning with structured knowledge.
Mental health symptoms predict flexible inference use
Study shows that specific mental health symptoms predict how people use flexible reasoning when making complex inferences, with alterations in goal-directed decision-making as a key mechanism.
Forensic knowledge depends on managing transformation, uncertainty, and warrant
Theoretical analysis of forensic knowledge production addressing technical instrumentalism through ontology, transformation processes, and epistemic capture mechanisms in institutional structures.
Maximal effort linked to greater preference for correct performance
Active inference model reveals voluntary mental effort is governed by motivation for accuracy rather than inhibition of habitual responses in Stroop task performance.
Liger+ dynamically balances latency and throughput in large model inference
Distributed inference system using interleaved parallelism to dynamically balance latency-throughput trade-offs via task-aware batch management and strategic kernel scheduling across multiple GPUs.
