Research
On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

168,742 papers · 148 categories

Trend · papers per month

56112167223 · Jun 202019922001200920172026
48 results for adversarial prompting

Diffusion LLMs can efficiently generate harmful prompts for adversarial testing.

problem Generating harmful prompts for adversarial testing is resource-intensive and costly.
method Transformed adversarial prompt optimization into an efficient inference task using pretrained Diffusion LLMs.
result Only a few conditional samples are required to generate harmful prompts with high reward.

ADAPT improves robustness of Vision Transformers without full model fine-tuning.

problem Vulnerability of Vision Transformers to adversarial attacks.
method Parameter-efficient prompt tuning with ADAPT framework for adaptive adversarial training.
result ADAPT achieves robust accuracy of ~40% w.r.t. SOTA methods using only ~1% of the parameters.

Shorter adversarial prompts help protect LLMs from jailbreak attacks.

problem Protecting large language models from jailbreak attacks with long adversarial suffixes.
method Adversarial training on shorter adversarial suffixes to defend against longer adversarial suffixes.
result Aligning LLMs on shorter adversarial suffixes can effectively defend against jailbreak attacks with longer suffixes.

VERA uses variational inference to jailbreak LLMs without manual optimization.

problem Lack of principled objective for gradient-based optimization in jailbreaking LLMs.
method VERA casts black-box jailbreak prompting as a variational inference problem, training a small attacker LLM to approximate the target LLM's posterior over adversarial prompts.
result VERA achieves strong performance across various target LLMs, demonstrating the value of probabilistic inference for adversarial prompt generation.

Gradient-based framework for optimizing text prompts in diffusion models.

problem Efficiently optimizing prompts in text-to-image diffusion models with large domain space and non-differentiable embeddings.
method Formulated as discrete optimization over language space, designed compact subspaces, and introduced shortcut text gradient.
result Empirically discovered prompts that enhance or destroy image faithfulness.

SmoothLLM defends LLMs from jailbreaking attacks by randomly perturbing inputs.

problem Adversaries can fool large language models into generating objectionable content.
method SmoothLLM randomly perturbs multiple copies of a prompt and aggregates predictions to detect adversarial inputs.
result SmoothLLM sets the state-of-the-art for robustness against various jailbreak attacks.

AdvMind detects adversary intent in black-box attacks with high accuracy.

problem Detecting adversary intent in black-box adversarial attacks is challenging.
method AdvMind accounts for adversary adaptiveness and synthesizes queries to expose intent.
result AdvMind detects adversary intent with over 75% accuracy after observing less than 3 query batches.

This paper improves continuous adversarial training for LLMs using in-context learning theory.

problem Efficiently defending large language models (LLMs) against jailbreak attacks.
method The paper presents a theoretical analysis of continuous adversarial training (CAT) for LLMs based on in-context learning (ICL) theory, proving a robust generalization bound and proposing an improved regularization term.
result The robust generalization bound explains why CAT can defend against jailbreak prompts and shows that LLM robustness is related to embedding matrix singular values.

NegToMe uses images to guide text-based models away from unwanted visual elements.

problem Insufficient text-based adversarial guidance for complex visual concepts.
method Negative token merging (NegToMe) using visual features from reference images.
result Significantly enhances output diversity and reduces visual similarity to copyrighted content.

New framework improves interpretability of trainable prompts.

problem Improving task-specific LLM performance with soft prompts remains a black-box method.
method Developed a theoretical framework for evaluating interpretability of trainable prompts, inspired new objective functions.
result Found a fundamental trade-off between interpretability and task performance in trainable prompts.

Study on how attention in prompt-tuning affects large language models.

problem Limited theoretical understanding of prompt-tuning and attention in LLMs.
method Exploration of prompt-tuning for one-layer attention architectures, contextual mixture-models, and self-contained prompt-attention model.
result Softmax-prompt-attention is more expressive than self-attention and linear-prompt-attention under contextual data model.

InfoPrompt improves soft prompt tuning by maximizing mutual information, leading to better performance.

problem High sensitivity of prompt tuning to initial conditions and insufficient task-relevant information.
method Develops an information-theoretic framework to maximize mutual information between prompts and model parameters, using novel loss functions.
result InfoPrompt accelerates convergence and outperforms traditional methods.

Understanding optimal prompts for binary sequence predictors is challenging.

problem Finding good prompts for binary sequence predictors is difficult.
method Viewing prompting as finding the best conditioning sequence on a near-optimal sequence predictor, using empirical and statistical analysis.
result Optimal prompts can be better understood given the pretraining distribution, which is not usually available.

Transformer models can approximate smooth functions with prompts, enhancing LLMs' dynamic capabilities.

problem Lack of theoretical framework for prompt engineering in transformer models.
method Formal framework demonstrating transformer models can approximate ββ-times differentiable functions with prompts.
result Transformer models can approximate ββ-times differentiable functions with arbitrary precision using appropriately structured prompts.

Single-head transformers with a single self-attention layer can approximate any sequence-to-sequence function and are efficient under certain conditions.

problem Statistical and computational limits of prompt tuning for transformer-based models.
method Investigation of single-head transformers with a single self-attention layer, proving universality and efficiency under SETH.
result Existence of almost-linear time prompt tuning inference algorithms under certain conditions.

TRIPLE efficiently optimizes prompts with a budget constraint.

problem Efficiently selecting good prompts from a pool of candidates.
method TRIPLE connects prompt optimization to best arm identification in MAB, leveraging BAI-FB tools.
result TRIPLE outperforms baselines on multiple tasks with limited budget constraints.

VERA-V uses variational inference to discover vulnerabilities in multimodal vision-language models.

problem Existing methods for jailbreaking vision-language models are brittle, limited, and focus on single attacks.
method VERA-V recasts jailbreak discovery as learning a joint posterior distribution over text-image prompts, using variational inference and three complementary strategies.
result VERA-V consistently outperforms state-of-the-art baselines, achieving up to 53.75% higher attack success rate.

Simple adaptive attacks can jailbreak state-of-the-art safety-aligned language models.

problem Vulnerability of safety-aligned language models to simple adaptive attacks.
method Designing adversarial prompt templates, random search on suffixes, and transfer/prefilling attacks.
result 100% attack success rate on various models including GPT-4, Vicuna, Mistral, and Claude.

Reward collapse occurs when ranking-based reward models yield uniform rewards for different prompts.

problem Reward collapse in aligning large language models with human preferences.
method Introduced a prompt-aware optimization scheme to derive closed-form expressions for reward distributions.
result Our prompt-aware utility functions significantly alleviate reward collapse during training.

eva method improves RL models by adaptively creating prompts, boosting performance.

problem Fixed prompt distribution limits scalability of RL models post-training.
method Evolving Alignment via Asymmetric Self-Play (eva) approach.
result Significant performance boost on benchmarks, e.g. 51.6% to 60.1% win-rate.

New framework tackles deep financial reporting bottleneck by improving hallucination and coherence.

problem Statistical smoothing trap in LLMs limits deep financial reporting quality.
method DeepNews Framework integrates information foraging, schema-guided planning, and adversarial prompting.
result DeepNews system achieves 25% acceptance rate in blind test, significantly outperforming SOTA.

The paper studies estimation rates for MoE models with a new prompt.

problem Estimating parameters in a softmax-contaminated MoE model.
method Analytic notion of distinguishability, minimax optimal estimation rates.
result Estimation rates are minimax optimal under distinguishability, but slower otherwise.

LanguaShrink compresses prompts using psycholinguistic principles to reduce costs.

problem Reducing computational cost and efficiency in large language model inference.
method Leverages psycholinguistic principles and the Ebbinghaus memory curve to compress prompts.
result Achieves up to 26 times compression while maintaining semantic similarity.

Develops PromptShift-CRC for drift-aware conformal risk control in foundation models under prompt and domain shift.

problem Fixed calibration risk in foundation models due to prompt and domain shift.
method Embeds prompts and responses, measures drift, gives more weight to recent examples, and updates risk online.
result Develops method to control risk up to terms for distribution mismatch and weighted quantile uncertainty.

ZegOT uses optimal transport to zero-shot segment images with text prompts.

problem Zero-shot semantic segmentation with limited image-text alignment knowledge.
method ZegOT uses optimal transport to match multiple text prompts with frozen image embeddings.
result ZegOT achieves state-of-the-art performance in zero-shot semantic segmentation.

This work analyzes CoT prompting methods from a statistical estimation perspective.

problem Improving the effectiveness of LLMs in solving multi-step reasoning problems.
method Introducing a multi-step latent variable model to characterize CoT prompting from a statistical estimation viewpoint.
result The CoT estimator is equivalent to a Bayesian estimator when the pretraining dataset is large.

Transformers can emulate various algorithms by prompting, proving universality.

problem How to emulate algorithms using fixed-weight Transformers.
method Two modes of in-context algorithm emulation: task-specific and prompt-programmable. Constructing prompts that encode algorithm parameters into token representations.
result Fixed-weight Transformers can emulate a broad class of algorithms via prompts.

OTSeg uses multi-prompt Sinkhorn attention to improve zero-shot semantic segmentation.

problem Leveraging pre-trained CLIP knowledge to align text embeddings with pixel embeddings.
method OTSeg employs Multi-Prompts Sinkhorn (MPS) and Multi-Prompts Sinkhorn Attention (MPSA) to enhance semantic feature matching.
result OTSeg achieves state-of-the-art performance in zero-shot semantic segmentation tasks.

Study analyzes convergence of parameter estimation in contaminated mixture of experts.

problem Challenges in learning from prompts in large-scale models.
method Convergence analysis, distinguishability condition, partial differential equations.
result Comprehensive convergence rates and minimax lower bounds for parameter estimation.

A method to improve LLMs by automating the construction of a mixture of expert prompts.

problem Limitation of single instruction prompts in covering complex problem spaces.
method Divide the problem space into sub-regions, each governed by a specialized expert with both an instruction and demos. A two-phase process constructs these experts.
result Achieves an average win rate of 81% across major benchmarks.

CurveRL optimizes large model reasoning by reweighting prompts based on their rank and density.

problem Improving large language model reasoning through context reweighting.
method Formulated prompt reweighting as a functional derivative, proposing CurveRL based on quantile coordinate transform.
result CurveRL consistently outperforms existing methods across multiple benchmarks.

Efficiently evaluate generative models at the prompt level using tensor factorization.

problem Fine-grained evaluations of generative models are costly and often misaligned with human judgment.
method Tensor factorization model that merges cheap autorater data with a small set of human gold-standard labels.
result The method provides accurate and tight confidence intervals for model performance.

Co-PLNet combines point and line predictions to improve wireframe parsing accuracy and efficiency.

problem Separate line and point predictions lead to inconsistent wireframes.
method Co-PLNet uses a Point-Line Prompt Encoder to convert early point detections into spatial prompts, which guide line refinement.
result Co-PLNet achieves better accuracy and robustness in wireframe parsing compared to existing methods.

Heavy-tailed outliers are more resilient to robust estimation than adversarial ones.

problem Developing robust estimators for data with outliers.
method Analyzing the relationship between adversarial and heavy-tailed outlier models.
result Optimal estimators for heavy-tailed outliers are also optimal for adversarial settings, but not vice versa.