COVERCAL selects PTQ calibration samples via weighted set cover over outlier channels, with a stylized clipping model showing missed coverage upper-bounds surrogate loss, yielding gains over random and other baselines on LLaMA and Mistral models.
Measuring massive multitask language understanding
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
citation-role summary
dataset 1
citation-polarity summary
years
2026 5roles
dataset 1polarities
use dataset 1representative citing papers
A causal audit via neuron-row zeroing shows attribution methods (LRP, IG) faithfully identify dispensable neurons and can install refusal behavior, while rank-stability proxies systematically miss selector failures.
SAFESEAL is a key-conditioned LLM watermarking framework using tournament sampling for synonym substitution and a contrastive detector that reports 98.2% detection, 0.983 BERTScore, and 0.963 entity similarity while claiming robustness to attacks.
DEL is a new loss for LLM numerical learning that applies supervised digit entropy optimization and extends to floating-point numbers, showing improved accuracy and distance metrics over prior methods on math benchmarks.
ORCE decouples answer generation from confidence estimation in LLMs and applies rank-based reinforcement learning on sampled completions to better align verbalized confidence with actual correctness likelihood.
citing papers explorer
-
Coverage-Based Calibration for Post-Training Quantization via Weighted Set Cover over Outlier Channels
COVERCAL selects PTQ calibration samples via weighted set cover over outlier channels, with a stylized clipping model showing missed coverage upper-bounds surrogate loss, yielding gains over random and other baselines on LLaMA and Mistral models.
-
Faithfulness to Refusal: A Causal Audit of Neuron Selectors
A causal audit via neuron-row zeroing shows attribution methods (LRP, IG) faithfully identify dispensable neurons and can install refusal behavior, while rank-stability proxies systematically miss selector failures.
-
Robust LLM Watermarking with Minimal Semantic Distortion for IP Protection
SAFESEAL is a key-conditioned LLM watermarking framework using tournament sampling for synonym substitution and a contrastive detector that reports 98.2% detection, 0.983 BERTScore, and 0.963 entity similarity while claiming robustness to attacks.
-
DEL: Digit Entropy Loss for Numerical Learning of Large Language Models
DEL is a new loss for LLM numerical learning that applies supervised digit entropy optimization and extends to floating-point numbers, showing improved accuracy and distance metrics over prior methods on math benchmarks.
-
ORCE: Order-Aware Alignment of Verbalized Confidence in Large Language Models
ORCE decouples answer generation from confidence estimation in LLMs and applies rank-based reinforcement learning on sampled completions to better align verbalized confidence with actual correctness likelihood.