ensemble-methods

Tag

Cards List
#ensemble-methods

Counterfactual Graph for Multi-Agent LLM Calibration

arXiv cs.CL · 3d ago Cached

This paper introduces CAGE, a counterfactual graph-based method for calibrating multi-agent LLM systems, evaluating on benchmarks like TriviaQA and MMLU-Pro across various communication topologies. The method outperforms existing post-hoc and LLM-elicited calibration approaches.

0 favorites 0 likes
#ensemble-methods

Linear Ensembles Wash Away Watermarks: On the Fragility of Distributional Perturbations in LLMs

arXiv cs.CL · 3d ago Cached

This paper reveals a fundamental vulnerability in LLM watermarking: when users have access to multiple models, averaging their output distributions cancels watermark perturbations, enabling detection evasion. The authors propose WASH and demonstrate empirically that averaging 3-5 models suppresses detection z-scores below thresholds while improving text quality.

0 favorites 0 likes
#ensemble-methods

Comparative Evaluation of Machine Learning Approaches for Minority-Class Financial Distress Prediction Under Class Imbalance Constraints

arXiv cs.LG · 2026-05-15 Cached

This paper presents a comparative evaluation of classical, ensemble, and neural machine learning approaches for predicting financial distress under severe class imbalance, using SMOTE for oversampling and SHAP for interpretability.

0 favorites 0 likes
#ensemble-methods

Pitfalls of Unlabeled Disagreement-Based Drift Detection in Streaming Tree Ensembles

arXiv cs.LG · 2026-05-14 Cached

This paper investigates disagreement-based drift detection in ensembles of incremental decision trees, finding that while effective in neural networks, the method underperforms loss-based detectors for tree ensembles due to limited model plasticity.

0 favorites 0 likes
#ensemble-methods

UCB exploration via Q-ensembles

OpenAI Blog · 2017-06-05 Cached

OpenAI presents a novel exploration strategy for deep reinforcement learning using ensembles of Q-functions with upper-confidence bounds (UCB), demonstrating significant performance improvements on the Atari benchmark.

0 favorites 0 likes
#ensemble-methods

Semi-supervised knowledge transfer for deep learning from private training data

OpenAI Blog · 2016-10-18 Cached

OpenAI presents PATE (Private Aggregation of Teacher Ensembles), a privacy-preserving approach that trains a student model on noisy outputs from multiple teacher models trained on disjoint datasets, providing strong differential privacy guarantees without exposing sensitive training data.

0 favorites 0 likes
← Back to home

Submit Feedback