Spend Experts Where You Are Unsure: Confidence-Adaptive Routing for Mixture-of-Experts LoRA

• URL: http://arxiv.org/abs/2607.26052v2
• Date: Sun, 02 Aug 2026 21:52:05 GMT
• Title: Spend Experts Where You Are Unsure: Confidence-Adaptive Routing for Mixture-of-Experts LoRA
• Abstract summary: We introduce CARE (Confidence-Adaptive Routing of Experts), which admits experts in a nucleus fashion.<n> CARE improves over fixed top-k MoE-LoRA at matched compute and matches the fixed-k=4 baseline while activating fewer experts.<n>The same confidence and disagreement signals also improve out-of-distribution detection over MSP, entropy, and multi-pass proxies.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Mixture-of-Experts (MoE) variants of Low-Rank Adaptation (LoRA) route every token to a fixed number of experts $k$. Tokens differ in how uncertain the model is about them, so a single k over-spends on easy tokens and under-serves hard ones. We observe that the router's output distribution is already a per-token uncertainty signal: peaked mass indicates confidence, while a flat distribution indicates ambiguity. We introduce CARE (Confidence-Adaptive Routing of Experts), which admits experts in a nucleus fashion. Experts are activated in decreasing router weight until their cumulative mass reaches a threshold, with a small extension when the admitted experts disagree. A budget thermostat calibrates the threshold so that the average number of active experts matches any target. CARE is a drop-in, single-forward-pass rule with no extra parameters. Across eight commonsense benchmarks on LLaMA-3.1-8B and Qwen2.5-7B, as well as math, code, and knowledge tasks, CARE improves over fixed top-k MoE-LoRA at matched compute and matches the fixed-k=4 baseline while activating fewer experts. The same confidence and disagreement signals also improve out-of-distribution detection over MSP, entropy, and multi-pass proxies. We support the design with nucleus fidelity, budget optimality, and an epistemic reading of disagreement, and we release code.

Related papers

• RAPTOR: Role-Aware Private Training for Mixture-of-Experts [68.5626338219096]
  RAPTOR is a Role-Aware Private Training framework.<n>RAPTOR alternates shared and expert optimization and targets each failure directly.<n> Experiments on Switch Transformer and OLMoE fine-tuning across GLUE tasks, and on DeepSeek-VL2-Tiny, show consistent gains over standard DP baselines.
  arXiv  Detail & Related papers  (2026-09-04T23:16:12Z)
• From Positionwise Confidence to Prefix Scheduling: Verifier Skipping in Speculative Decoding [41.41611499810411]
  Speculative diffusion decoding (SDD) generates every position in a draft block in parallel with a discrete diffusion model.<n>This paper recognizes that this creates a new control handle: whether to invoke the verifier at all.
  arXiv  Detail & Related papers  (2026-08-14T18:00:08Z)
• Share First, Route What Remains: A Unified Framework for Token-Adaptive MoE Computation [49.8459545291965]
  Mixture-of-experts (MoE) models have recently moved beyond routing a fixed number of complete experts.<n>We study this dependency by decomposing sparsely upcycled feed-forward experts into key-value channels.<n>We instantiate it in UniF-MoE, a unified framework for token-adaptive MoE computation.
  arXiv  Detail & Related papers  (2026-08-11T02:40:58Z)
• Reward-Free Evolving Agents via Pairwise Validator [12.6053820984668]
  A self-evolving agentic loop repeatedly proposes a tweaked version of an agent.<n>It accepts or rejects the change based on a per-iteration quality signal.<n>We propose replacing the scalar at the accept/reject gate with a pairwise validator.
  arXiv  Detail & Related papers  (2026-07-15T22:50:32Z)
• Best-Arm Identification with Generative Proxy [10.674965991983974]
  We study fixed-confidence best-arm identification in which each costly reward pull is paired with a cheap but correlated proxy score.<n>We propose PROBE, a phase-elimination algorithm that maintains an upper certificate on the residual variance with an ordinary least squares fit.<n>We prove that PROBE is $$-PAC and attains the known-correlation oracle sample complexity up to a constant multiplicative factor and a constant additive calibration cost.
  arXiv  Detail & Related papers  (2026-07-08T00:41:43Z)
• ARCA: Adapter-Residual Credit Assignment When Token Signals Degenerate [0.0]
  Token-level credit assignment for language-model reinforcement learning is usually formulated as if the policy were fully trainable.<n>We propose a lightweight alternative that derives token salience from the adapter's own hidden-state residual.<n>In a compact MATH/Qwen3-1.7B GRPO sweep, ARCA exhibits the predicted non-degenerate middle-regime credit distribution under matched rollout budgets.
  arXiv  Detail & Related papers  (2026-05-29T18:42:06Z)
• Unleashing Implicit Rewards: Prefix-Value Learning for Distribution-Level Optimization [74.91418266859297]
  Implicit process reward models (PRMs) provide fine-grained reward signals along the reasoning process.<n>PRMs suffer from a train-inference mismatch: training only constrains a sequence-level aggregate, whereas inference requires token-level scores to reflect local step quality.<n>We propose a novel Implicit Prefix-Value Reward Model (IPVRM), which directly learns a prefix-conditioned value function estimating the probability of eventual correctness.<n>We also propose Distribution-Level RL (DistRL), which computes TD advantages for both sampled tokens and high-probability candidate tokens.
  arXiv  Detail & Related papers  (2026-04-14T18:19:54Z)
• Locally Confident, Globally Stuck: The Quality-Exploration Dilemma in Diffusion Language Models [52.61023005303122]
  We show that low-confidence remasking improves a proxy for quality while constraining the entropy of the induced sequence distribution.<n>We develop a simple Independent--Hastings sampler that approximately targets this distribution during decoding.
  arXiv  Detail & Related papers  (2026-04-01T02:01:30Z)
• Do LLMs Know They Are Being Tested? Evaluation Awareness and Incentive-Sensitive Failures in GPT-OSS-20B [1.948261185683419]
  We investigate whether "evaluation scent" inflates measured performance without commensurate capability gains.<n>We run six paired A/B scenarios that hold task content and decoding fixed while varying framing.<n>We provide a reproducible A/B framework (prompt banks, validators, per-run scores, scripts) and practical guidance.
  arXiv  Detail & Related papers  (2025-10-08T09:49:05Z)
• Sample Smart, Not Hard: Correctness-First Decoding for Better Reasoning in LLMs [72.82403830490084]
  We argue that the decoding rule should be calibrated by correctness, not confidence alone.<n>We propose simple strategies that achieve this goal: Greedy-Threshold makes sampling greedy at very low confidence steps.<n>Together, our findings challenge prevailings about decoding under uncertainty and show gains across math and general reasoning benchmarks.
  arXiv  Detail & Related papers  (2025-10-07T14:46:12Z)
• From Score Distributions to Balance: Plug-and-Play Mixture-of-Experts Routing [52.01745035243826]
  Mixture-of-Experts (MoE) models can scale parameter capacity by routing each token to a subset of experts.<n> conditional routing shifts the burden on inference memory, limiting the number of experts per device.<n>We present LASER, a plug-and-play, inference-time routing algorithm that balances load while preserving accuracy.
  arXiv  Detail & Related papers  (2025-09-29T16:29:17Z)
• Continuous K-Max Bandits [54.21533414838677]
  We study the $K$-Max multi-armed bandits problem with continuous outcome distributions and weak value-index feedback.<n>This setting captures critical applications in recommendation systems, distributed computing, server scheduling, etc.<n>Our key contribution is the computationally efficient algorithm DCK-UCB, which combines adaptive discretization with bias-corrected confidence bounds.
  arXiv  Detail & Related papers  (2025-02-19T06:37:37Z)
• Patch-level Routing in Mixture-of-Experts is Provably Sample-efficient for Convolutional Neural Networks [74.68583356645276]
  In deep learning, mixture-of-experts (MoE) activates one or few experts (sub-networks) on a per-sample or per-token basis. We show for the first time that pMoE provably reduces the required number of training samples to achieve desirable generalization.
  arXiv  Detail & Related papers  (2023-06-07T00:16:10Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.