arXiv subjects

T. Ben Thompson

Publications and source records attributed to T. Ben Thompson.

4 recordsLinked to original sources

arXiv2024

Guarantees for Comprehensive Simulation Assessment of Statistical Methods

Simulation can evaluate a statistical method for properties such as Type I Error, FDR, or bias on a grid of hypothesized parameter values. But what about the gaps between the grid-points? Continuous Simulation Extension (CSE) is a proof-by-simulation framework which can supplement simulations with (1) confidence bands valid over regions of parameter space or (2) calibration of rejection thresholds to provide rigorous proof of strong Type I Error control. CSE extends simulation estimates at grid-points into bounds over nearby space using a model shift bound related to the Renyi divergence, which we analyze for models in exponential family or canonical GLM form. CSE can work with adaptive sampling, nuisance parameters, administrative censoring, multiple arms, multiple testing, Bayesian randomization, Bayesian decision-making, and inference algorithms of arbitrary complexity. As a case study, we calibrate for strong Type I Error control a Phase II/III Bayesian selection design with 4 unknown statistical parameters. Potential applications include calibration of new statistical procedures or streamlining regulatory review of adaptive trial designs. Our open-source software implementation imprint is available athttps://github.com/Confirm-Solutions/imprint

stat.ME↗

arXiv2024

FLRT: Fluent Student-Teacher Redteaming

Many publicly available language models have been safety tuned to reduce the likelihood of toxic or liability-inducing text. To redteam or jailbreak these models for compliance with toxic requests, users and security analysts have developed adversarial prompting techniques. One attack method is to apply discrete optimization techniques to the prompt. However, the resulting attack strings are often gibberish text, easily filtered by defenders due to high measured perplexity, and may fail for unseen tasks and/or well-tuned models. In this work, we improve existing algorithms (primarily GCG and BEAST) to develop powerful and fluent attacks on safety-tuned models like Llama-2 and Phi-3. Our technique centers around a new distillation-based approach that encourages the victim model to emulate a toxified finetune, either in terms of output probabilities or internal activations. To encourage human-fluent attacks, we add a multi-model perplexity penalty and a repetition penalty to the objective. We also enhance optimizer strength by allowing token insertions, token swaps, and token deletions and by using longer attack sequences. The resulting process is able to reliably jailbreak the most difficult target models with prompts that appear similar to human-written prompts. On Advbench we achieve attack success rates $>93$% for Llama-2-7B, Llama-3-8B, and Vicuna-7B, while maintaining model-measured perplexity $<33$; we achieve $95$% attack success for Phi-3, though with higher perplexity. We also find a universally-optimized single fluent prompt that induces $>88$% compliance on previously unseen tasks across Llama-2-7B, Phi-3-mini and Vicuna-7B and transfers to other black-box models.

cs.CL↗

arXiv2024

Fluent dreaming for language models

Feature visualization, also known as "dreaming", offers insights into vision models by optimizing the inputs to maximize a neuron's activation or other internal component. However, dreaming has not been successfully applied to language models because the input space is discrete. We extend Greedy Coordinate Gradient, a method from the language model adversarial attack literature, to design the Evolutionary Prompt Optimization (EPO) algorithm. EPO optimizes the input prompt to simultaneously maximize the Pareto frontier between a chosen internal feature and prompt fluency, enabling fluent dreaming for language models. We demonstrate dreaming with neurons, output logits and arbitrary directions in activation space. We measure the fluency of the resulting prompts and compare language model dreaming with max-activating dataset examples. Critically, fluent dreaming allows automatically exploring the behavior of model internals in reaction to mildly out-of-distribution prompts. Code for running EPO is available at https://github.com/Confirm-Solutions/dreamy. A companion page demonstrating code usage is at https://confirmlabs.org/posts/dreamy.html

cs.CL↗

arXiv2017

Enabling large-scale viscoelastic calculations via neural network acceleration

One of the most significant challenges involved in efforts to understand the effects of repeated earthquake cycle activity are the computational costs of large-scale viscoelastic earthquake cycle models. Computationally intensive viscoelastic codes must be evaluated thousands of times and locations, and as a result, studies tend to adopt a few fixed rheological structures and model geometries, and examine the predicted time-dependent deformation over short (<10 yr) time periods at a given depth after a large earthquake. Training a deep neural network to learn a computationally efficient representation of viscoelastic solutions, at any time, location, and for a large range of rheological structures, allows these calculations to be done quickly and reliably, with high spatial and temporal resolution. We demonstrate that this machine learning approach accelerates viscoelastic calculations by more than 50,000%. This magnitude of acceleration will enable the modeling of geometrically complex faults over thousands of earthquake cycles across wider ranges of model parameters and at larger spatial and temporal scales than have been previously possible.

physics.geo-ph↗