In the paper, we consider multilevel multitask analysis of cryptocurrency news using a fine-tuned Mistral 7B large language model with retrieval-augmented generation (RAG). On the first level of analytics, the fine-tuned model generates graph and text summaries with sentiment scores as well as JSON representations of summaries. Higher levels perform hierarchical stacking that consolidates sets of graph-based and text-based summaries as well as summaries of summaries into comprehensive reports. The combination of graph and text summaries provides complementary views of cryptocurrency news. The model is fine-tuned with 4-bit quantization using the PEFT/LoRA approach. The representation of cryptocurrency news as knowledge graph can essentially eliminate problems with large language model hallucinations. The obtained results demonstrate that the use of fine-tuned Mistral 7B LLM models for multilevel cryptocurrency news analysis can conduct informative qualitative and quantitative analytics, providing important insights.

Relation extraction enables the construction of structured knowledge for many downstream applications. While large language models (LLMs) have shown great promise in this domain, most existing methods concentrate on relation classification, which predicts the semantic relation type between a related entity pair. However, we observe that LLMs often struggle to reliably determine whether a relation exists, especially in cases involving complex sentence structures or intricate semantics, which leads to spurious predictions. Such hallucinations can introduce noisy edges in knowledge graphs, compromising the integrity of structured knowledge and downstream reliability. To address these challenges, we propose DEPTH, a framework that integrates Dependency-aware sEntence simPlification and Two-tiered Hierarchical refinement into the relation extraction pipeline. Given a sentence and its candidate entity pairs, DEPTH operates in two stages: (1) the Grounding module extracts relations for each pair by leveraging their shortest dependency path, distilling the sentence into a minimal yet coherent relational context that reduces syntactic noise while preserving key semantics; (2) the Refinement module aggregates all local predictions and revises them based on a holistic understanding of the sentence, correcting omissions and inconsistencies. We further introduce a causality-driven reward model that mitigates reward hacking by disentangling spurious correlations, enabling robust fine-tuning via reinforcement learning with human feedback. Experiments on six benchmarks demonstrate that DEPTH reduces the average hallucination rate to 7.0\% while achieving a 17.2\% improvement in average F1 score over state-of-the-art baselines.

Data annotation refers to the labeling or tagging of textual data with relevant information. A large body of works have reported positive results on leveraging LLMs as an alternative to human annotators. However, existing studies focus on classic NLP tasks, and the extent to which LLMs as data annotators perform in domains requiring expert knowledge remains underexplored. In this work, we investigate comprehensive approaches across three highly specialized domains and discuss practical suggestions from a cost-effectiveness perspective. To the best of our knowledge, we present the first systematic evaluation of LLMs as expert-level data annotators.

Generative Large language models (LLMs) have demonstrated remarkable capabilities for a wide range of applications, but reducing ungrounded or erroneous responses remains a major growth area. Unlike task-specific models, there lack an effective method to calibrate the confidence level of LLM responses to indicate potential errors and facilitate human-in-the-loop verification. An important source of calibration stems from expert-stipulated programmatic supervision, which is often available at low cost but has its own limitations such as noise and coverage. In this paper, we introduce a Pareto optimal self-supervision framework that can leverage available programmatic supervision to systematically calibrate LLM responses by producing a risk score for every LLM response, without any additional manual efforts. This is accomplished by learning a harmonizer model to align with LLM output as well as other weak supervision sources. The model assigns higher risk scores to more uncertain LLM responses and facilitate error correction. Experiments on standard relation extraction and classification tasks in biomedical and general domains demonstrate that the proposed risk score is highly correlated with the actual LLM error rate. By using a dynamic prompting strategy based on the risk score, we observed significant accuracy improvement for off-the-shelf LLMs, boosting GPT-3.5 results past state-of-the-art (SOTA) weak supervision model and GPT-4 results past SOTA supervised results on challenging evaluation datasets.