The surge in applications of large language models (LLMs) has prompted concerns about the generation of misleading or fabricated information, known as hallucinations. Therefore, detecting hallucinations has become critical to maintaining trust in LLM-generated content. A primary challenge in learning a truthfulness classifier is the lack of a large amount of labeled truthful and hallucinated data. To address the challenge, we introduce HaloScope, a novel learning framework that leverages the unlabeled LLM generations in the wild for hallucination detection. Such unlabeled data arises freely upon deploying LLMs in the open world, and consists of both truthful and hallucinated information. To harness the unlabeled data, we present an automated membership estimation score for distinguishing between truthful and untruthful generations within unlabeled mixture data, thereby enabling the training of a binary truthfulness classifier on top. Importantly, our framework does not require extra data collection and human annotations, offering strong flexibility and practicality for real-world applications. Extensive experiments show that HaloScope can achieve superior hallucination detection performance, outperforming the competitive rivals by a significant margin.

Preference learning algorithms (e.g., RLHF and DPO) are frequently used to steer LLMs to produce generations that are more preferred by humans, but our understanding of their inner workings is still limited.

Environment Climate Change Antarctica has a'gravity hole' The geological oddity has existed since dinosaurs roamed the Earth. Breakthroughs, discoveries, and DIY tips sent six days a week. A "gravity hole" beneath Antarctica sounds like the plot to a bad sci-fi movie, but it's a very real situation deep beneath the Earth's surface stretching back tens of millions of years. The phenomenon thankfully isn't as apocalyptic as it sounds, either. In fact, researchers say these complex interactions between rock densities, gravitational pull, and sea levels are actually helping them understand how the southernmost continent's ice sheets evolved, and what their influences mean for the planet's climate.

Diffusion models are powerful, but they require a lot of time and data to train. We propose Patch Diffusion, a generic patch-wise training framework, to significantly reduce the training time costs while improving data efficiency, which thus helps democratize diffusion model training to broader users. At the core of our innovations is a new conditional score function at the patch level, where the patch location in the original image is included as additional coordinate channels, while the patch size is randomized and diversified throughout training to encode the cross-region dependency at multiple scales. Sampling with our method is as easy as in the original diffusion model.

Limitations One limitation of our model is its potential for data bias. This could limit the applications of the model. MLLMs could be used to create fake news articles or social media posts. Hyperparameters Number of layers 24 Hidden size 2,048 FFN inner hidden size 8,192 Attention heads 32 Dropout 0.1 Attention dropout 0.1 Activation function GeLU [1] V ocabulary size 64,007 Soft tokens V size 64 Max length 2,048 Relative position embedding xPos [2] Initialization Magneto [3] Table 1: Hyperparameters of causal language model of K The detailed instruction tuning hyperparameters are listed in Table 3. The models are trained on web-scale multimodal corpora.