Natural Language Processing

CulturePark, a novel multi-agent communication framework, generates high-quality cross-cultural data to fine-tune LLMs, significantly reducing cultural bias and boosting cross-cultural understanding.

CultureLLM, a new approach, effectively incorporates cultural nuances into LLMs using semantic data augmentation, significantly outperforming existing models.

One-time training improves multiple LLMs using a tiny portable model, drastically reducing costs and resource needs for model enhancement.

CRITICEVAL: A new benchmark reliably evaluates LLMs’ ability to identify and correct flaws in their responses, addressing limitations of existing methods by offering comprehensive and reliable evaluat…

LLM-based text embeddings are powerful but lack interpretability. This paper introduces QA-Emb, a novel method that uses an LLM to answer yes/no questions about a text, thereby producing an interpreta…

CoVoMix: Generating human-like, multi-speaker conversations with zero-shot speech synthesis.

Coke: A cost-efficient KBQA strategy using LLMs and KGMs, maximizing accuracy while minimizing GPT-4 fees by up to 20.89%

CorDA: Context-oriented weight decomposition enhances large language model fine-tuning by integrating task context, improving efficiency and mitigating catastrophic forgetting.

LLMs struggle to cooperate sustainably; GOVSIM reveals this, showing communication and ‘universalization’ reasoning improve outcomes.

Researchers developed a novel continual learning method achieving state-of-the-art performance by aligning data representations across tasks using pre-trained tokens, eliminating the need for experien…

CONTEXTCITE pinpoints which parts of a given context led a language model to generate a specific statement, improving model verification and response quality.

Researchers created a Materials Knowledge Graph (MKG) using large language models to efficiently organize and integrate knowledge from a decade of high-quality materials science research, enhancing da…

ConStat: Exposing hidden LLM contamination!

LLMs surprisingly infer censored knowledge from implicit training data hints, posing safety challenges.

Conformal Alignment certifies trustworthy foundation model outputs by guaranteeing a user-specified fraction meet alignment criteria, regardless of the model or data.

LLMs regulate uncertainty via specialized ’entropy’ and ’token frequency’ neurons, impacting prediction confidence without directly altering logits.

Boost language model performance across domains with ‘Concentration’: a new prompt optimization objective that prioritizes stable, deep-layer attention.

CALDERA: a new post-training LLM compression algorithm achieving state-of-the-art zero-shot performance using low-rank, low-precision decomposition.

LLM-directed compositional 3D-aware video generation (C3V) achieves high-fidelity video generation with diverse motion and flexible concept control by decomposing prompts, generating 3D concepts, and …

MINITRON: Efficiently creating smaller, high-performing LLMs via pruning & distillation, slashing training costs by up to 40x!