The emerging capabilities of large language models (LLMs) have sparked concerns about their immediate potential for harmful misuse.The core approach to mitigate these concerns is the detection of harmful queries to the model.Current detection approaches are fallible, and are particularly susceptible to attacks that exploit mismatched generalization of model capabilities (e.g., prompts in lowresource languages or prompts provided in non-text modalities such as image and audio).To tackle this challenge, we propose OMNI-GUARD, an approach for detecting harmful prompts across languages and modalities.Our approach (i) identifies internal representations of an LLM/MLLM that are aligned across languages or modalities and then (ii) uses them to build a language-agnostic or modality-agnostic classifier for detecting harmful prompts.OM-NIGUARD improves harmful prompt classification accuracy by 11.57% over the strongest baseline in a multilingual setting, by 20.44% for image-based prompts, and sets a new SOTA for audio-based prompts.By repurposing embeddings computed during generation, OMNI-GUARD is also very efficient (≈ 120× faster than the next fastest baseline).Code and data are available at https://github.com/ vsahil/OmniGuard.

In the past few years, Language Models (LMs) have shown par-human capabilities in several domains. Despite their practical applications and exceeding user consumption, they are susceptible to jailbreaks when malicious input exploits the LM’s weaknesses, causing it to deviate from its intended behavior. Current defensive strategies either classify the input prompt as adversarial or prevent LMs from generating harmful outputs. However, it is challenging to explain the reason behind the malicious nature of the jailbreak, which results in a wide variety of closed-box approaches. In this research, we propose and demonstrate that system-prompt attention from Small Language Models (SLMs) can be used to characterize adversarial prompts, providing a novel, explainable, and cheaper defense approach called AttentionDefense. Our research suggests that the attention mechanism is an integral component in understanding and explaining how LMs respond to malicious input that is not captured in the semantic meaning of text embeddings. The proposed AttentionDefense is evaluated against existing jailbreak benchmark datasets. Ablation studies show that SLM-based AttentionDefense has equivalent or better jailbreak detection performance compared to text embedding-based classifiers and GPT-4 zero-shot detectors. To further validate the efficacy of the proposed approach, we generate a dataset of novel jailbreak variants of the existing benchmark dataset using a closed-loop LLM-based multi-agent system. We demonstrate that the proposed AttentionDefense approach performs robustly on this novel jailbreak dataset while existing approaches suffer in performance. Additionally, for practical purposes AttentionDefense is an ideal solution as it has the computation requirements of a small LM but the performance of a LLM detector.

Benchmarks have emerged as the central approach for evaluating Large Language Models (LLMs).The research community often relies on a model's average performance across the test prompts of a benchmark to evaluate the model's performance.This is consistent with the assumption that the test prompts within a benchmark represent a random sample from a real-world distribution of interest.We note that this is generally not the case; instead, we hold that the distribution of interest varies according to the specific use case.We find that (1) the correlation in model performance across test prompts is non-random, (2) accounting for correlations across test prompts can change model rankings on major benchmarks, (3) explanatory factors for these correlations include semantic similarity and common LLM failure points.

Bioinformatics (2016) 32(5), 690–696. doi:10.1093/bioinformatics/btv633 The authors of the above paper wish to inform readers that there was a typographical error in Equations 2 and 6 of the published paper. The corrected equations are given below. The paper has now been corrected online.

MOTIVATION: Current differential correlation methods are designed to determine molecular feature pairs that have the largest magnitude of difference between correlation coefficients. These methods do not easily capture molecular feature pairs that experience no correlation in one group but correlation in another, which may reflect certain types of biological interactions. We have developed a tool, the Discordant method, which categorizes the correlation types for each group to make this possible. RESULTS: We compare the Discordant method to existing approaches using simulations and two biological datasets with different types of -omics data. In contrast to other methods, Discordant identifies phenotype-related features at a similar or higher rate while maintaining reasonable computational tractability and usability. AVAILABILITY AND IMPLEMENTATION: R code and sample data are available at https://github.com/siskac/discordant CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.