Precision in Practice: Knowledge Guided Code Summarizing Grounded in Industrial Expectations

Code summaries are essential for helping developers understand code functionality and reducing maintenance and collaboration costs. Although recent advances in large language models (LLMs) have significantly improved automatic code summarization, the practical usefulness of generated summaries in industrial settings remains insufficiently explored. In collaboration with documentation experts from the industrial HarmonyOS project, we conducted a questionnaire study showing that over 57.4% of code summaries produced by state-of-the-art approaches were rejected due to violations of developers' expectations for industrial documentation. Beyond semantic similarity to reference summaries, developers emphasize additional requirements, including the use of appropriate domain terminology, explicit function categorization, and the avoidance of redundant implementation details. To address these expectations, we propose ExpSum, an expectation-aware code summarization approach that integrates function metadata abstraction, informative metadata filtering, context-aware domain knowledge retrieval, and constraint-driven prompting to guide LLMs in generating structured, expectation-aligned summaries. We evaluate ExpSum on the HarmonyOS project and widely used code summarization benchmarks. Experimental results show that ExpSum consistently outperforms all baselines, achieving improvements of up to 26.71% in BLEU-4 and 20.10% in ROUGE-L on HarmonyOS. Furthermore, LLM-based evaluations indicate that ExpSum-generated summaries better align with developer expectations across other projects, demonstrating its effectiveness for industrial code documentation.

MemCast: Memory-Driven Time Series Forecasting with Experience-Conditioned Reasoning

Time series forecasting (TSF) plays a critical role in decision-making for many real-world applications. Recently, LLM-based forecasters have made promising advancements. Despite their effectiveness, existing methods often lack explicit experience accumulation and continual evolution. In this work, we propose MemCast, a learning-to-memory framework that reformulates TSF as an experience-conditioned reasoning task. Specifically, we learn experience from the training set and organize it into a hierarchical memory. This is achieved by summarizing prediction results into historical patterns, distilling inference trajectories into reasoning wisdom, and inducing extracted temporal features into general laws. Furthermore, during inference, we leverage historical patterns to guide the reasoning process and utilize reasoning wisdom to select better trajectories, while general laws serve as criteria for reflective iteration. Additionally, to enable continual evolution, we design a dynamic confidence adaptation strategy that updates the confidence of individual entries without leaking the test set distribution. Extensive experiments on multiple datasets demonstrate that MemCast consistently outperforms previous methods, validating the effectiveness of our approach. Our code is available at https://github.com/Xiaoyu-Tao/MemCast-TS.

HALT: Hallucination Assessment via Log-probs as Time series

Hallucinations remain a major obstacle for large language models (LLMs), especially in safety-critical domains. We present HALT (Hallucination Assessment via Log-probs as Time series), a lightweight hallucination detector that leverages only the top-20 token log-probabilities from LLM generations as a time series. HALT uses a gated recurrent unit model combined with entropy-based features to learn model calibration bias, providing an extremely efficient alternative to large encoders. Unlike white-box approaches, HALT does not require access to hidden states or attention maps, relying only on output log-probabilities. Unlike black-box approaches, it operates on log-probs rather than surface-form text, which enables stronger domain generalization and compatibility with proprietary LLMs without requiring access to internal weights. To benchmark performance, we introduce HUB (Hallucination detection Unified Benchmark), which consolidates prior datasets into ten capabilities covering both reasoning tasks (Algorithmic, Commonsense, Mathematical, Symbolic, Code Generation) and general purpose skills (Chat, Data-to-Text, Question Answering, Summarization, World Knowledge). While being 30x smaller, HALT outperforms Lettuce, a fine-tuned modernBERT-base encoder, achieving a 60x speedup gain on HUB. HALT and HUB together establish an effective framework for hallucination detection across diverse LLM capabilities.

CoMeT: Collaborative Memory Transformer for Efficient Long Context Modeling

The quadratic complexity and indefinitely growing key-value (KV) cache of standard Transformers pose a major barrier to long-context processing. To overcome this, we introduce the Collaborative Memory Transformer (CoMeT), a novel architecture that enables LLMs to handle arbitrarily long sequences with constant memory usage and linear time complexity. Designed as an efficient, plug-in module, CoMeT can be integrated into pre-trained models with only minimal fine-tuning. It operates on sequential data chunks, using a dual-memory system to manage context: a temporary memory on a FIFO queue for recent events, and a global memory with a gated update rule for long-range dependencies. These memories then act as a dynamic soft prompt for the next chunk. To enable efficient fine-tuning on extremely long contexts, we introduce a novel layer-level pipeline parallelism strategy. The effectiveness of our approach is remarkable: a model equipped with CoMeT and fine-tuned on 32k contexts can accurately retrieve a passkey from any position within a 1M token sequence. On the SCROLLS benchmark, CoMeT surpasses other efficient methods and achieves performance comparable to a full-attention baseline on summarization tasks. Its practical effectiveness is further validated on real-world agent and user behavior QA tasks. The code is available at: https://anonymous.4open.science/r/comet-B00B/

Can Vision-Language Models Handle Long-Context Code? An Empirical Study on Visual Compression

Large Language Models (LLMs) struggle with long-context code due to window limitations. Existing textual code compression methods mitigate this via selective filtering but often disrupt dependency closure, causing semantic fragmentation. To address this, we introduce LongCodeOCR, a visual compression framework that renders code into compressed two-dimensional image sequences for Vision-Language Models (VLMs). By preserving a global view, this approach avoids the dependency breakage inherent in filtering. We systematically evaluate LongCodeOCR against the state-of-the-art LongCodeZip across four benchmarks spanning code summarization, code question answering, and code completion. Our results demonstrate that visual code compression serves as a viable alternative for tasks requiring global understanding. At comparable compression ratios ($\sim$1.7$\times$), LongCodeOCR improves CompScore on Long Module Summarization by 36.85 points over LongCodeZip. At a 1M-token context length with Glyph (a specialized 9B VLM), LongCodeOCR maintains higher accuracy than LongCodeZip while operating at about 4$\times$ higher compression. Moreover, compared with LongCodeZip, LongCodeOCR drastically reduces compression-stage overhead (reducing latency from $\sim$4.3 hours to $\sim$1 minute at 1M tokens). Finally, our results characterize a fundamental coverage--fidelity trade-off: visual code compression retains broader context coverage to support global dependencies, yet faces fidelity bottlenecks on exactness-critical tasks; by contrast, textual code compression preserves symbol-level precision while sacrificing structural coverage.

ContextEvolve: Multi-Agent Context Compression for Systems Code Optimization

Large language models are transforming systems research by automating the discovery of performance-critical algorithms for computer systems. Despite plausible codes generated by LLMs, producing solutions that meet the stringent correctness and performance requirements of systems demands iterative optimization. Test-time reinforcement learning offers high search efficiency but requires parameter updates infeasible under API-only access, while existing training-free evolutionary methods suffer from inefficient context utilization and undirected search. We introduce ContextEvolve, a multi-agent framework that achieves RL-level search efficiency under strict parameter-blind constraints by decomposing optimization context into three orthogonal dimensions: a Summarizer Agent condenses semantic state via code-to-language abstraction, a Navigator Agent distills optimization direction from trajectory analysis, and a Sampler Agent curates experience distribution through prioritized exemplar retrieval. This orchestration forms a functional isomorphism with RL-mapping to state representation, policy gradient, and experience replay-enabling principled optimization in a textual latent space. On the ADRS benchmark, ContextEvolve outperforms state-of-the-art baselines by 33.3% while reducing token consumption by 29.0%. Codes for our work are released at https://anonymous.4open.science/r/ContextEvolve-ACC

From Latent Signals to Reflection Behavior: Tracing Meta-Cognitive Activation Trajectory in R1-Style LLMs

R1-style LLMs have attracted growing attention for their capacity for self-reflection, yet the internal mechanisms underlying such behavior remain unclear. To bridge this gap, we anchor on the onset of reflection behavior and trace its layer-wise activation trajectory. Using the logit lens to read out token-level semantics, we uncover a structured progression: (i) Latent-control layers, where an approximate linear direction encodes the semantics of thinking budget; (ii) Semantic-pivot layers, where discourse-level cues, including turning-point and summarization cues, surface and dominate the probability mass; and (iii) Behavior-overt layers, where the likelihood of reflection-behavior tokens begins to rise until they become highly likely to be sampled. Moreover, our targeted interventions uncover a causal chain across these stages: prompt-level semantics modulate the projection of activations along latent-control directions, thereby inducing competition between turning-point and summarization cues in semantic-pivot layers, which in turn regulates the sampling likelihood of reflection-behavior tokens in behavior-overt layers. Collectively, our findings suggest a human-like meta-cognitive process-progressing from latent monitoring, to discourse-level regulation, and to finally overt self-reflection. Our analysis code can be found at https://github.com/DYR1/S3-CoT.

Internal Flow Signatures for Self-Checking and Refinement in LLMs

Large language models can generate fluent answers that are unfaithful to the provided context, while many safeguards rely on external verification or a separate judge after generation. We introduce \emph{internal flow signatures} that audit decision formation from depthwise dynamics at a fixed inter-block monitoring boundary. The method stabilizes token-wise motion via bias-centered monitoring, then summarizes trajectories in compact \emph{moving} readout-aligned subspaces constructed from the top token and its close competitors within each depth window. Neighboring window frames are aligned by an orthogonal transport, yielding depth-comparable transported step lengths, turning angles, and subspace drift summaries that are invariant to within-window basis choices. A lightweight GRU validator trained on these signatures performs self-checking without modifying the base model. Beyond detection, the validator localizes a culprit depth event and enables a targeted refinement: the model rolls back to the culprit token and clamps an abnormal transported step at the identified block while preserving the orthogonal residual. The resulting pipeline provides actionable localization and low-overhead self-checking from internal decision dynamics. \emph{Code is available at} \texttt{github.com/EavnJeong/Internal-Flow-Signatures-for-Self-Checking-and-Refinement-in-LLMs}.

Multi-task Code LLMs: Data Mix or Model Merge?

Recent research advocates deploying smaller, specialized code LLMs in agentic frameworks alongside frontier models, sparking interest in efficient strategies for multi-task learning that balance performance, constraints, and costs. We compare two approaches for creating small, multi-task code LLMs: data mixing versus model merging. We conduct extensive experiments across two model families (Qwen Coder and DeepSeek Coder) at two scales (2B and 7B parameters), fine-tuning them for code generation and code summarization tasks. Our evaluation on HumanEval, MBPP, and CodeXGlue benchmarks reveals that model merging achieves the best overall performance at larger scale across model families, retaining 96% of specialized model performance on code generation tasks while maintaining summarization capabilities. Notably, merged models can even surpass individually fine-tuned models, with our best configuration of Qwen Coder 2.5 7B model achieving 92.7% Pass@1 on HumanEval compared to 90.9% for its task-specific fine-tuned equivalent. At a smaller scale we find instead data mixing to be a preferred strategy. We further introduce a weight analysis technique to understand how different tasks affect model parameters and their implications for merging strategies. The results suggest that careful merging and mixing strategies can effectively combine task-specific capabilities without significant performance degradation, making them ideal for resource-constrained deployment scenarios.

DART-ing Through the Drift: Dynamic Tracing of Knowledge Neurons for Adaptive Inference-Time Pruning

Large Language Models (LLMs) exhibit substantial parameter redundancy, particularly in Feed-Forward Networks (FFNs). Existing pruning methods suffer from two primary limitations. First, reliance on dataset-specific calibration introduces significant data dependency and computational overhead. Second, being predominantly static, they fail to account for the evolving subset of knowledge neurons in LLMs during autoregressive generation as the context evolves. To address this, we introduce DART, i.e., Dynamic Attention-Guided Runtime Tracing), a lightweight, training-free method that performs on-the-fly context-based pruning. DART monitors shifts in attention score distributions to infer context changes, dynamically updating neuron-level masks to retain salient parameters. Across ten benchmarks, DART outperforms prior dynamic baseline, achieving accuracy gains of up to 14.5% on LLAMA-3.1-8B at 70% FFN sparsity. Furthermore, DART achieves up to 3x better ROUGE-L scores with respect to static-masked pruning on summarization tasks, with its performance comparable to the original dense models. We conclusively demonstrate that the proposed framework effectively adapts to diverse semantic contexts, preserves model capabilities across both general and domain-specific tasks while running at less than 10MBs of memory for LLAMA-3.1-8B(16GBs) with 0.1% FLOPs overhead. The code is available at https://github.com/seeder-research/DART.