Related papers: Beyond Inference-Only Deployment: Comparing Weight…

Do all instances need inference through the big models for a correct prediction? Perhaps not; some instances are easy and can be answered correctly by even small capacity models. This provides opportunities for improving the computational…

Large language models (LLMs) have demonstrated strong reasoning capabilities, and as existing approaches for enhancing LLM reasoning continue to mature, increasing attention has shifted toward meta-reasoning as a promising direction for…

Standard LLM cascades improve efficiency by deferring difficult queries from weak to strong models. However, these systems are typically static: when faced with repeated or semantically similar queries, they redundantly consult the…

Modern key-value storage engines built on Log-Structured Merge-trees (LSM-trees), such as RocksDB and LevelDB, rely heavily on the performance of their compaction operations, which are impacted by a complex set of interdependent…

Scaling model size, training data, and compute power have driven advances in large language models (LLMs), but these approaches are reaching saturation as human-generated text is exhausted and further gains diminish. We propose experience…

Large language models (LLMs) increasingly receive information as streams of passages, conversations, and long-context workflows. While longer context windows expose more evidence, they do not ensure that useful information is preserved and…

In-context learning (ICL) enables large language models (LLMs) to adapt to new tasks without weight updates by learning from demonstration sequences. While ICL shows strong empirical performance, its internal representational mechanisms are…

Scaling test-time compute via long Chain-of-Thought unlocks remarkable gains in reasoning capabilities, yet it faces practical limits due to the linear growth of KV cache and quadratic attention complexity. In this paper, we introduce…

As reasoning LLMs increasingly trade tokens for accuracy through deliberation, search, and self-correction, a single accuracy score can no longer tell whether those tokens buy useful reasoning, recovery from hard instances, or unnecessary…

Vision-Language Models (VLMs) are expensive because the LLM processes hundreds of largely redundant visual tokens. Existing token reduction methods typically exploit \textit{either} vision-encoder saliency (broad but query-agnostic)…

Large language models (LLMs) have gained much attention in the recommendation community; some studies have observed that LLMs, fine-tuned by the cross-entropy loss with a full softmax, could achieve state-of-the-art performance already.…

Large language models (LLMs) have shown remarkable promise but remain challenging to continually improve through traditional finetuning, particularly when integrating capabilities from other specialized LLMs. Popular methods like ensemble…

Scientific reasoning rarely stops at what is directly observable; it often requires uncovering hidden structure from data. From estimating reaction constants in chemistry to inferring demand elasticities in economics, this latent structure…

Progress in LLMs is increasingly measured through standardized benchmarks, where state-of-the-art improvements are often separated by fractions of a percentage point. At the same time, the computational cost of evaluating modern LLMs has…

As the knowledge of large language models (LLMs) becomes outdated over time, there is a growing need for efficient methods to update them, especially when injecting proprietary information. Our study reveals that comprehension-intensive…

Embedding fusion has emerged as an effective approach for enhancing performance across various NLP tasks. However, systematic guidelines for selecting optimal layers and developing effective fusion strategies for the integration of LLMs…

In Continual Learning (CL), a model is required to learn a stream of tasks sequentially without significant performance degradation on previously learned tasks. Current approaches fail for a long sequence of tasks from diverse domains and…

The organization of latent knowledge within large-scale models poses unique challenges when addressing overlapping representations and optimizing contextual accuracy. Conceptual redundancies embedded across layers often result in…

Large language models (LLMs) continue to struggle with knowledge-intensive questions that require up-to-date information and multi-hop reasoning. Augmenting LLMs with hybrid external knowledge, such as unstructured text and structured…

Large language models are increasingly deployed in settings where reliability matters, yet output-level uncertainty signals such as token probabilities, entropy, and self-consistency can become brittle under calibration--deployment…