4arXiv cs.CL (Computation and Language)·
KnowledgeDebugger: GUI tool for knowledge localization and editing in Transformers
Researchers introduce KnowledgeDebugger, a no-code GUI tool for exploring how Transformers store and process knowledge and for applying knowledge editing methods. The tool wraps EasyEdit, a widely-used knowledge editing library, and is inspired by LM-Debugger, providing interactive access to state-of-the-art editing approaches without requiring code. It targets the exploratory phase of knowledge editing research, enabling case-study-level investigation before committing to large-scale experiments.
Related guides (2)
Related events (8)
KATE framework improves LLM tool calling via experiential knowledge integration and parallel reasoning
Researchers present KATE (Knowledge-Augmented Tool Execution), a framework addressing LLM failures in multi-step tool use by systematically studying knowledge acquisition, activation, and internalization. Key findings include that instance-level experiential knowledge outperforms abstract intent-level knowledge, that expanding reasoning width via parallel sampling with aggregation beats deeper chain-of-thought, and that reinforcement learning outperforms supervised fine-tuning for knowledge internalization. KATE is evaluated on BFCL-V3 and AppWorld benchmarks, showing consistent improvements over strong baselines across model scales.
Understand-Anything: interactive knowledge graph tool for code exploration using AI assistants
Egonex-AI has released Understand-Anything, a TypeScript tool that converts codebases into interactive knowledge graphs that can be explored, searched, and queried. The tool integrates with multiple AI coding assistants including Claude Code, Codex, Cursor, GitHub Copilot, and Gemini CLI. It has accumulated 62,256 GitHub stars with 1,146 added today, indicating strong community traction.
Knowledge editing via locate-then-edit transferred to masked diffusion language models, revealing multi-token failure mode
A new arXiv paper investigates whether locate-then-edit knowledge editing methods, developed for autoregressive models, transfer to masked diffusion language models (MDMs) such as LLaDA and Dream. The authors find that causal tracing identifies the same early-to-mid-layer MLP location in both paradigms, but MDMs degrade systematically on multi-token edits due to partially unmasked intermediate states that the edit was never optimized for. A correction targeting these intermediate states substantially restores multi-token editing performance. The work is the first systematic comparison of knowledge editing across autoregressive and diffusion-based language model paradigms.
CodeGraph: Pre-indexed Local Code Knowledge Graph for AI Coding Agents
CodeGraph is an open-source TypeScript tool that builds a pre-indexed knowledge graph of a codebase to reduce token usage and tool calls for AI coding agents including Claude Code, Codex, Cursor, OpenCode, and Hermes Agent. It runs entirely locally, positioning itself as an efficiency layer between codebases and LLM-based coding assistants. The project gained significant traction with 3,688 stars in a single day, reaching 16,371 total stars.
Dual-adapter routing system improves knowledge editing precision in LLMs
A new arXiv paper introduces a route-specialized dual-adapter architecture for knowledge editing in LLMs, separating the concerns of writing edits (edit adapter) and suppressing them when irrelevant (locality adapter). A relevance router gates which adapter is applied, addressing the locality problem in memory-assisted editing. Evaluated on CounterFact, zsRE, and MQuAKE benchmarks using Llama-3.1-8B-Instruct and Qwen3-8B, the method achieves best-in-class probability-preference accuracy across all three datasets. Ablations show the gain comes from the architectural separation rather than increased parameter capacity.
code-review-graph: Local Knowledge Graph for Claude Code with 6.8×–49× Token Reduction
A community-built Python tool that constructs a persistent local knowledge graph of a codebase to optimize Claude Code's context usage. Claims 6.8× token reduction on code reviews and up to 49× on daily coding tasks by feeding the model only relevant code sections. The project has accumulated 16,880 stars with 123 added today, indicating strong community traction.
HumP-KD: Uncertainty-aware multi-stage knowledge distillation for efficient fire classification
Researchers propose HumP-KD, a knowledge distillation framework that compresses two heterogeneous transformer teachers (Swin-Tiny and ViT-Base) into a lightweight MobileViT-S student for real-time fire classification. The student model achieves 0.9876 mean F1 on a 31K-image dataset while retaining only 4.94M parameters—a 5.7× reduction over Swin-Tiny—and runs at 37.72 CPU FPS. The framework combines hierarchical feature alignment, spatial attention masking, and progressive multi-stage distillation to maintain accuracy under degraded visual conditions.
Graphify: open-source tool to convert codebases into queryable knowledge graphs for AI coding assistants
Graphify is a Python library that converts arbitrary code folders, SQL schemas, scripts, docs, and media into a queryable knowledge graph, designed to serve as a skill or context layer for AI coding assistants including Claude Code, Codex, Cursor, and Gemini CLI. The project has accumulated 72,438 GitHub stars with 504 added today, indicating strong community traction. It targets the problem of giving AI coding agents unified, structured access to heterogeneous project artifacts across code, schema, and infrastructure.