arXiv:2608. 15154v1 Announce Type: cross Abstract: Population-based adaptive importance sampling (AIS) methods use a set of proposal densities to approximate complex target distributions.
arXiv:2609.24969v1 Announce Type: new Abstract: As agents are deployed with increased autonomy, even extremely rare events along their stochastic output trajectories can occur and prove catastrophic....
The paper introduces FAMIS, a flow-based multiple importance sampling framework that learns a nonuniform mixture of normalizing flow proposals for rare‑event estimation. It does not need presampled failure data or prior knowledge of failure modes, instead adapting the mixture through sequential evaluations of the limit state function. The method employs a smooth rare‑event surrogate, a tempered target sequence, defensive exploration, Rao‑Blackwellized weight updates, and a Jensen‑Shannon repulsion term to promote diversity, achieving accurate failure probability estimates with fewer training samples and stable variance reduction in complex reliability problems.
Chopthin-Consensus Power Sampling (CCPS) is a new inference-time decoding method for large language models that uses the Chopthin resampler to preserve diversity among particle trajectories. By enforcing an upper bound on weight ratios instead of equal-weight resampling, CCPS maintains a richer set of distinct reasoning paths and guarantees a lower bound on effective sample size. Coupled with a semantic-majority selection mechanism, CCPS achieves higher oracle coverage and matches or surpasses baseline accuracy on multiple reasoning benchmarks.
arXiv:2606. 08460v1 Announce Type: cross Abstract: Data-adaptive two-sample testing assesses if two samples come from the same distribution, using a discrepancy learned from the data (e.
arXiv:2606. 20820v2 Announce Type: replace Abstract: Can we trust evaluation scores to capture an LLM's true real-world performance?
arXiv:2607. 16240v1 Announce Type: cross Abstract: Direct Alignment Algorithms (DAAs) such as DPO have become a common way to post-train and align LLMs with human preferences.
arXiv:2506. 01486v2 Announce Type: replace Abstract: Data imbalance persists as a pervasive challenge in regression tasks, introducing bias in model performance and undermining predictive reliability.
arXiv:2607. 04728v1 Announce Type: cross Abstract: Reinforcement learning (RL) post-training for large language models (LLMs) follows a efficient paradigm of "rollout then update", which inevitably results in off-policy training data.
arXiv:2502. 11027v5 Announce Type: replace Abstract: Large language model (LLM) scaling inference is key to unlocking greater performance, and leveraging diversity has proven an effective way to enhance it.
The paper introduces Balance of Benchmarks (BoB), a framework that improves task-conditioned model comparison by weighting benchmark evidence based on semantic density, equating scores across varying difficulty levels, and pooling task-relevant residuals. BoB retains all eligible benchmark data while adjusting its influence, outperforming uniform averaging on the WildScores dataset with higher Spearman correlation, lower MAE, and better shortlist hit rates. The method also reduces ranking instability when benchmarks are repeated or paraphrased, and lowers retrospective regret in model selection.
The paper investigates how AI systems that perform best‑of‑n search require different validation strategies as the search width changes. It shows that auditing only small search widths leaves a gap in reliability estimates for larger widths, and proposes retaining candidate ranks and truth labels to estimate reliability across all widths up to N. The authors derive theoretical bounds on the minimax mean‑squared error, design procedures that achieve these bounds, and demonstrate that a shared audit can significantly reduce maximum error across many widths in practical CodeRM pools.
arXiv:2606. 27242v1 Announce Type: new Abstract: Training-free source selection for LLM families with shared vocabularies arises in scientific string domains such as SMILES, protein, and genomic sequences, where candidate corpora share a tokenizer but differ in prediction targets.