Student-Guided Teacher Distillation for Efficient LLM Task Routing: Positioning Against Jev-Style System-1 Classifiers

arXiv:2610.02516 · cs.LG, cs.AI · Submitted 2026-10-01 · Read on arXiv

cs.LG, cs.AI

Submitted: 2026-10-01

Updated: 2026-10-01

Comments: 13 pages, 2 figures, 1 table

Code: https://github.com/NandhaKishorM/laya

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Distillation — 1× in this paper · explained in 2 episodes

Sources

• RLCD: Reinforcement Learning from Contrastive Distillation for Language Model Alignment
• Distilling System 2 into System 1
• Distilling the Knowledge in a Neural Network
• Dense Passage Retrieval for Open-Domain Question Answering
• Confident Adaptive Language Modeling
• Fast Inference from Transformers via Speculative Decoding
• Accelerating Large Language Model Decoding with Speculative Sampling
• AttentionXML: Label Tree-based Attention-Aware Deep Model for High-Performance Extreme Multi-Label Text Classification
• DiSMEC - Distributed Sparse Machines for Extreme Multi-label Classification
• Taming Pretrained Transformers for Extreme Multi-label Text Classification
• Fast Multi-Resolution Transformer Fine-tuning for Extreme Multi-label Text Classification
• Benchmarking Zero-shot Text Classification: Datasets, Evaluation and Entailment Approach
• Building Efficient Universal Classifiers with Natural Language Inference
• Self-training with Noisy Student improves ImageNet classification
• Born Again Neural Networks
• Weak-to-Strong Generalization: Eliciting Strong Capabilities With Weak Supervision
• RouteLLM: Learning to Route LLMs with Preference Data
• FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performance
• Tryage: Real-time, intelligent Routing of User Prompts to Large Language Models
• RouterBench: A Benchmark for Multi-LLM Routing System

Related papers

• Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
• AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
• Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
• Convergence issues in Relational Concept Analysis based on AOC-posets
• Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks