…By temporally reversing real-world videos at zero cost as natural counterfactual samples, YoCausal establishes an arbitrarily extensible evaluation protocol. Level 1 introduces the Reverse Surprise Index (RSI), quantifying arrow-of-time…
…AI-generated summary Causal Transformer language models suffer from strictly sequential decoding and a quadratic per-step attention cost. While linear-time causal models and discrete diffusion models each address these weaknesses…
…Listed API prices vary by over an order of magnitude across providers, but we use price dispersion only as directional motivation, not as causal evidence of marginal cost. The core physical question…
…AI-generated summary Adapting pretrained models typically involves a trade-off between the high training costs of backpropagation and the heavy inference overhead of memory-based or in-context learning . We propose…
…Encoding cost stays roughly constant in K instead of scaling with it. Findings. Multi-token helps every diffusion backbone we test, on every benchmark (MS MARCO, TREC DL'19/'20, BEIR-7…
…Generated by Qwen/Qwen2.5-Coder-32B-Instruct Diffusion models have emerged as the backbone of modern generative AI, powering advances in vision, language, audio and other modalities. Despite their success, they…
…We introduce Policy Optimization with Internal State Value Estimation ), which obtains a baseline at negligible cost by using the policy model's internal signals already computed during the policy forward pass. A…
…AI-generated summary Mixture-of-Experts (MoE) architectures in Large Language Models (LLMs) have significantly reduced inference costs through sparse activation . However, this sparse activation paradigm also introduces new safety challenges. Since…
…Improving the Performance of Non-Thinking Models at No Cost (2026) Balanced Thinking: Improving Chain of Thought Training in Vision Language Models (2026) Scaling Reasoning Tokens via RL and Parallel Thinking: Evidence…
…All methods may achieve an estimated memory-bandwidth cost over 50% lower than BLT on generation tasks. Each approach offers its own unique advantages, together removing key barriers to the practical use…