Earlier schemes could only handle stochastic substitutions anddeletions, andthus weareaiming foramore natural and appealing robustness guarantee that holds with respect to editdistance.

We would like to express our sincere gratitude to the reviewers for providing their valuable feedback. This generalization will be added to the revision. We will clarify this point together with further experiments on purely real datasets in a revision. This can readily be obtained by [39, 40] which do not exploit the hierarchical structure. We will provide this discussion in a revision.

We show in this work that reinforcement learning can be successfully applied to decoding short to moderate length sparse graph-based channel codes. Specifically, we focus on low-density parity check (LDPC) codes, which for example have been standardized in the context of 5G cellular communication systems due to their excellent error correcting performance. These codes are typically decoded via belief propagation iterative decoding on the corresponding bipartite (Tanner) graph of the code via flooding, i.e., all check and variable nodes in the Tanner graph are updated at once. In contrast, in this paper we utilize a sequential update policy which selects the optimum check node (CN) scheduling in order to improve decoding performance. In particular, we model the CN update process as a multi-armed bandit process with dependent arms and employ a Q-learning scheme for optimizing the CN scheduling policy. In order to reduce the learning complexity, we propose a novel graph-induced CN clustering approach to partition the state space in such a way that dependencies between clusters are minimized. Our results show that compared to other decoding approaches from the literature, the proposed reinforcement learning scheme not only significantly improves the decoding performance, but also reduces the decoding complexity dramatically once the scheduling policy is learned.

The rapid evolution of Large Language Models (LLMs) has driven the need for increasingly sophisticated inference pipelines and hardware platforms. Modern LLM serving extends beyond traditional prefill-decode workflows, incorporating multi-stage processes such as Retrieval Augmented Generation (RAG), key-value (KV) cache retrieval, dynamic model routing, and multi step reasoning. These stages exhibit diverse computational demands, requiring distributed systems that integrate GPUs, ASICs, CPUs, and memory-centric architectures. However, existing simulators lack the fidelity to model these heterogeneous, multi-engine workflows, limiting their ability to inform architectural decisions. To address this gap, we introduce HERMES, a Heterogeneous Multi-stage LLM inference Execution Simulator. HERMES models diverse request stages; including RAG, KV retrieval, reasoning, prefill, and decode across complex hardware hierarchies. HERMES supports heterogeneous clients executing multiple models concurrently unlike prior frameworks while incorporating advanced batching strategies and multi-level memory hierarchies. By integrating real hardware traces with analytical modeling, HERMES captures critical trade-offs such as memory bandwidth contention, inter-cluster communication latency, and batching efficiency in hybrid CPU-accelerator deployments. Through case studies, we explore the impact of reasoning stages on end-to-end latency, optimal batching strategies for hybrid pipelines, and the architectural implications of remote KV cache retrieval. HERMES empowers system designers to navigate the evolving landscape of LLM inference, providing actionable insights into optimizing hardware-software co-design for next-generation AI workloads.

The authors are listed in alphabetical order.

The authors are listed in alphabetical order.

People with paralysis can now have their thoughts turned into speech just by imagining talking in their heads. While brain-computer interfaces can already decode the neural activity of people with paralysis when they physically attempt speaking, this can require a fair amount of effort. So Benyamin Meschede-Krasa at Stanford University and his colleagues sought a less energy-intensive approach. Take control of your brain's master switch to optimise how you think "We wanted to see whether there were similar patterns when someone was simply imagining speaking in their head," he says. "And we found that this could be an alternative, and indeed, a more comfortable way for people with paralysis to use that kind of system to restore their communication."

In human-AI collaboration, a central challenge is deciding whether the AI should handle a task, be deferred to a human expert, or be addressed through collaborative effort. Existing Learning to Defer approaches typically make binary choices between AI and humans, neglecting their complementary strengths. They also lack interpretability, a critical property in high-stakes scenarios where users must understand and, if necessary, correct the model's reasoning. To overcome these limitations, we propose Defer-and-Complement Decision-Making via Decoupled Concept Bottleneck Models (DeCoDe), a concept-driven framework for human-AI collaboration. DeCoDe makes strategy decisions based on human-interpretable concept representations, enhancing transparency throughout the decision process. It supports three flexible modes: autonomous AI prediction, deferral to humans, and human-AI collaborative complementarity, selected via a gating network that takes concept-level inputs and is trained using a novel surrogate loss that balances accuracy and human effort. This approach enables instance-specific, interpretable, and adaptive human-AI collaboration. Experiments on real-world datasets demonstrate that DeCoDe significantly outperforms AI-only, human-only, and traditional deferral baselines, while maintaining strong robustness and interpretability even under noisy expert annotations.

Reasoning in large language models (LLMs) tends to produce substantially longer token generation sequences than simpler language modeling tasks. This extended generation length reflects the multi-step, compositional nature of reasoning and is often correlated with higher solution accuracy. From an efficiency perspective, longer token generation exacerbates the inherently sequential and memory-bound decoding phase of LLMs. However, not all parts of this expensive reasoning process are equally difficult to generate. We leverage this observation by offloading only the most challenging parts of the reasoning process to a larger, more capable model, while performing most of the generation with a smaller, more efficient model; furthermore, we teach the smaller model to identify these difficult segments and independently trigger offloading when needed. To enable this behavior, we annotate difficult segments across 18k reasoning traces from the OpenR1-Math-220k chain-of-thought (CoT) dataset. We then apply supervised fine-tuning (SFT) and reinforcement learning fine-tuning (RLFT) to a 1.5B-parameter reasoning model, training it to learn to offload the most challenging parts of its own reasoning process to a larger model. This approach improves AIME24 reasoning accuracy by 24% and 28.3% while offloading 1.35% and 5% of the generated tokens respectively. We open-source our SplitReason model, data, code and logs.