Large language models (LLMs) empowered by chain-of-thought reasoning have achieved impressive accuracy on complex tasks but suffer from excessive inference costs and latency when applied uniformly to all problems. We propose SABER (Switchable and Balanced Training for Efficient LLM Reasoning), a reinforcement learning framework that endows LLMs with user‑controllable, token‑budgeted reasoning. SABER first profiles each training example’s base‑model thinking token usage and assigns it to one of the predefined budget tiers. During fine‑tuning, the model is guided by system prompts and length‑aware rewards to respect its assigned budget. In parallel, we incorporate no‑think examples to ensure the model remains reliable even when explicit reasoning is turned off. SABER further supports four discrete inference modes—NoThink, FastThink, CoreThink, and DeepThink, enabling flexible trade‑offs between latency and reasoning depth. Extensive evaluations on math reasoning (MATH, GSM8K), code generation (MBPP), and logical reasoning (LiveBench-Reasoning) demonstrate that SABER achieves high accuracy under tight budgets, graceful degradation, and effective cross-scale and cross‑domain generalization. In particular, SABER‑FastThink cuts reasoning length by 65.4% and yields a 3.6% accuracy gain compared with the base model on the MATH benchmark.

While large-scale language models (LLMs) have demonstrated remarkable capabilities in specific natural language processing (NLP) tasks, they may still lack proficiency compared to specialized models in certain domains, such as grammatical error correction (GEC). Drawing inspiration from the concept of curriculum learning, we have delved into refining LLMs into proficient GEC experts by devising effective curriculum learning (CL) strategies. In this paper, we introduce a novel approach, termed LLM-based curriculum learning, which capitalizes on the robust semantic comprehension and discriminative prowess inherent in LLMs to gauge the complexity of GEC training data. Unlike traditional curriculum learning techniques, our method closely mirrors human expert-designed curriculums. Leveraging the proposed LLM-based CL method, we sequentially select varying levels of curriculums ranging from easy to hard, and iteratively train and refine using the pretrianed T5 and LLaMA series models. Through rigorous testing and analysis across diverse benchmark assessments in English GEC, including the CoNLL14 test, BEA19 test, and BEA19 development sets, our approach showcases a significant performance boost over baseline models and conventional curriculum learning methodologies. Specifically, our method achieves a new state-of-the-art (SOTA) result on the CoNLL14 test set, with an F 0 . 5 score of 69.6. Additionally, on the BEA19 test set and BEA19 development set, our approach outperforms conventional curriculum learning methodologies by 1.0 and 0.3 F 0 . 5 points, respectively.

Data Fusion is the combination of a number of independent search results, relating to the same document collection, into a single result to be presented to the user. A number of probabilistic data fusion models have been shown to be effective in empirical studies. These typically attempt to estimate the probability that particular documents will be relevant, based on training data. However, little attempt has been made to gauge how the accuracy of these estimations affect fusion performance. The focus of this paper is twofold: firstly, that accurate estimation of the probability of relevance results in effective data fusion; and secondly, that an effective approximation of this probability can be made based on less training data that has previously been employed. This is based on the observation that the distribution of relevant documents follows a similar pattern in most high-quality result sets. Curve fitting suggests that this can be modelled by a simple function that is less complex than other models that have been proposed. The use of existing IR evaluation metrics is proposed as a substitution for probability calculations. Mean Average Precision is used to demonstrate the effectiveness of this approach, with evaluation results demonstrating competitive performance when compared with related algorithms with more onerous requirements for training data.