Input-Blind Controls Produce Substantial Oracle Headroom for Layer Programs in Multiple-Choice Evaluation
cs.LG, cs.CL, cs.SE
Submitted: 2026-10-07
Updated: 2026-10-07
Sources
- Mixture-of-Recursions: Learning Dynamic Recursive Depths for Adaptive Token-Level Computation
- MACRO: Markov Chain Routing of Transformer Layers
- Training-Free Looped Transformers
- What Does a Routing Oracle Measure Under Stochastic Decoding? Coupling, Scorer Choice, and Single-Commit Ceilings
- Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge
- Training Verifiers to Solve Math Word Problems
- Neural Thickets: Diverse Task Experts Are Dense Around Pretrained Weights
- The Llama 3 Herd of Models
- Dr.LLM: Dynamic Layer Routing in LLMs
- Reachability Is Not Realization: Tracing the Sources of LLM Benchmark Gains
- Skip a Layer or Loop it? Test-Time Depth Adaptation of Pretrained LLMs
- Skip a Layer or Loop It? Learning Program-of-Layers in LLMs
- Deployable Per-Instance Multi-Layer Activation Steering for Large Language Models
- Opportunity Is Not Realizability: Selection-Valid Diagnostics for Multi-LLM Routing
- Architectural Sampling: Test-Time Scaling via Computational Diversity in Frozen Vision-Language Models
- Programs-of-Layers in LLMs through the Lens of Cortical Areas
- More Programs or More Rolls? Separating Coverage from Specialization in LLM Harnesses
- Qwen3 Technical Report
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