Same-sample two-point Gaussian ZO-SGD achieves Õ(d/T) last-iterate suboptimality with probability 1−δ under conditional sub-Gaussian noise, with only logarithmic 1/δ dependence.
and Chen, Danqi and Arora, Sanjeev , month = jan, year =
7 Pith papers cite this work, alongside 36 external citations. Polarity classification is still indexing.
7
Pith papers citing it
36
external citations · external index
citation-role summary
background 2
citation-polarity summary
roles
background 2polarities
background 2representative citing papers
Introduces MM-Privacy dataset and evaluations showing MLLMs leak sensitive data from images in various tasks, highlighting task inconsistency effects.
Zeroth-order LLM fine-tuning is reframed as an inference workload and run on vLLM, yielding 2.34x-8.13x speedups on OPT models with comparable accuracy to standard LoZO.
REALISTA generates semantically coherent adversarial prompts via latent-space optimization over input-dependent editing directions, achieving stronger hallucination elicitation than prior realistic attacks on open-source and reasoning LLMs.
REGLU guides LoRA-based unlearning via representation subspaces and orthogonal regularization to outperform prior methods on forget-retain trade-off in LLM benchmarks.
Influence scoring can use only forward passes: CountSketch-compressed outer products of the LM-head residual and final hidden state give accurate attribution and valuation from 14M to 32B parameters.
A comprehensive survey of PEFT algorithms for large models, covering their performance, overhead, applications, and real-world system implementations.
citing papers explorer
-
High-Probability Last-Iterate Guarantees for Two-Point Gaussian Zeroth-Order Stochastic Gradient Descent
Same-sample two-point Gaussian ZO-SGD achieves Õ(d/T) last-iterate suboptimality with probability 1−δ under conditional sub-Gaussian noise, with only logarithmic 1/δ dependence.
-
Unveiling Privacy Risks in Multi-modal Large Language Models: Task-specific Vulnerabilities and Mitigation Challenges
Introduces MM-Privacy dataset and evaluations showing MLLMs leak sensitive data from images in various tasks, highlighting task inconsistency effects.
-
LLM Zeroth-Order Fine-Tuning is an Inference Workload
Zeroth-order LLM fine-tuning is reframed as an inference workload and run on vLLM, yielding 2.34x-8.13x speedups on OPT models with comparable accuracy to standard LoZO.
-
REALISTA: Realistic Latent Adversarial Attacks that Elicit LLM Hallucinations
REALISTA generates semantically coherent adversarial prompts via latent-space optimization over input-dependent editing directions, achieving stronger hallucination elicitation than prior realistic attacks on open-source and reasoning LLMs.
-
Representation-Guided Parameter-Efficient LLM Unlearning
REGLU guides LoRA-based unlearning via representation subspaces and orthogonal regularization to outperform prior methods on forget-retain trade-off in LLM benchmarks.
-
Sketching the Readout of Large Language Models for Scalable Data Attribution and Valuation
Influence scoring can use only forward passes: CountSketch-compressed outer products of the LM-head residual and final hidden state give accurate attribution and valuation from 14M to 32B parameters.
-
Parameter-Efficient Fine-Tuning for Large Models: A Comprehensive Survey
A comprehensive survey of PEFT algorithms for large models, covering their performance, overhead, applications, and real-world system implementations.