6arXiv cs.CL (Computation and Language)·
Danus: Fact-Graph Memory Orchestration for Multi-Agent Mathematical Reasoning
Researchers introduce Danus, an open-source orchestration system for research-level mathematical reasoning that uses a shared fact graph as a global memory mechanism. The system coordinates a main planning agent with parallel worker agents and a stateless verifier that gates claims before they enter the shared proof state. Danus is evaluated on six case studies spanning algebraic geometry, singularity theory, and combinatorics, demonstrating the ability to construct long, incremental proofs on open research problems. The work addresses a key bottleneck in scaling LLM-based math agents: maintaining coherent, verifiable shared state across parallel proof search.
Related guides (1)
Related events (8)
DeepMath: A Lightweight Math Reasoning Agent with smolagents
Hugging Face published a blog post introducing DeepMath, a lightweight mathematical reasoning agent built on the smolagents framework. The post demonstrates how to construct a capable math reasoning agent using small models and tool-use patterns. This represents a practical application of the agent-tool ecosystem for specialized reasoning tasks.
GitOfThoughts: Git-based agent memory substrate with sobering findings on memory utility for novel problems
Researchers introduce GitOfThoughts, a system that stores LLM reasoning trees as git repositories, enabling replayable, auditable, and mergeable agent memory at low engineering cost. Across five memory substrates (none, markdown, vector, graph, git), two benchmarks, and two model scales with pre-registered replications, the paper finds that no memory format reliably improves accuracy on novel problems. Memory only helps above a 'copyability threshold' (similarity >~0.8), where retrieved cases are near-duplicates of the current problem — and even then, the gain is answer retrieval rather than method transfer. The paper also documents a retracted result and refuted hypothesis, modeling a rigorous evaluation standard.
DABStep: Data Agent Benchmark for Multi-step Reasoning
Hugging Face introduces DABStep, a benchmark designed to evaluate data agents on multi-step reasoning tasks. The benchmark targets agentic systems that must perform complex, sequential data operations rather than single-step queries. It aims to fill a gap in evaluation tooling for realistic data analysis workflows involving tool use and chained reasoning.
MemDreamer: Hierarchical graph memory and agentic retrieval for long video understanding
MemDreamer is a plug-and-play framework that decouples perception and reasoning for long-video understanding by incrementally building a three-tier Hierarchical Graph Memory capturing spatiotemporal and causal relations. During inference, a reasoning model uses an Observation-Reason-Action loop with agentic tool-augmented retrieval to navigate the memory graph, constraining the context window to 2% of full-context ingestion while achieving a 12.5-point absolute accuracy gain. The system reaches SOTA on four benchmarks, narrowing the gap with human experts to 3.7 points. The authors also report a strong linear correlation between logical reasoning performance and long-video understanding, proposing agentic capability scaling as a new paradigm for multimodal comprehension.
StreamMA: Streaming communication in multi-agent reasoning reduces latency and improves accuracy
Researchers introduce StreamMA, a multi-agent reasoning system that streams individual reasoning steps to downstream agents as they are generated, rather than waiting for a complete chain. This pipelining approach reduces end-to-end latency and also improves accuracy by shielding downstream agents from error-prone late reasoning steps. Evaluated across eight benchmarks, two frontier LLMs (Claude Opus 4.6 and GPT-5.4), and three topologies, StreamMA outperforms serial and single-agent baselines by an average of 7.3 percentage points. The paper also identifies a 'step-level scaling law' — a new scaling dimension orthogonal to agent-count scaling.
Goedel-Architect achieves state-of-the-art formal theorem proving with blueprint-based agentic framework
Goedel-Architect is an agentic framework for formal theorem proving in Lean 4 that uses blueprint generation — a dependency graph of definitions and lemmas — rather than recursive decomposition, enabling parallel lemma closure and global refinement. Built on DeepSeek-V4-Flash (284B-A13B), it achieves 99.2% pass@1 on MiniF2F-test and 75.6% on PutnamBench, scaling to 100% on MiniF2F, 88.8% on PutnamBench, and 4/6 on IMO 2025 when seeded with natural-language proofs. The authors claim state-of-the-art performance for an open-source pipeline at up to 500x lower cost than comparable systems.
REAL: Reasoning-enhanced temporal graph framework for LLM long-term memory management
REAL is a new framework that represents LLM conversational memory as a temporal, confidence-aware directed property graph, where atomic facts carry validity intervals, confidence scores, and exploration intent labels. It addresses three limitations of prior memory systems: flat text structures, destructive overwrites of evolving facts, and passive retrieval. The system uses non-destructive temporal updates, semantic evaluator-guided hybrid beam search, and counterfactual inference to repair incomplete retrieval states. Experiments show a 22.72% average improvement over flat-text, graph-based, and existing memory baselines.
VADAOrchestra: Neurosymbolic framework combining LLM orchestration with Datalog+/- symbolic reasoning for adaptive workflows
VADAOrchestra is a neurosymbolic framework that hybridizes LLM-based orchestration with a symbolic Datalog+/- inference engine to model complex, adaptive workflows. An LLM incrementally plans and encodes workflow steps as logic programs, while a dedicated symbolic engine handles all inference, decoupling orchestration from execution. The approach targets auditability, scalability over large datasets, and explainability — limitations of pure LLM-agent architectures — and is evaluated on real-world financial use cases. The work positions itself as a bridge between traditional Business Process Management rigidity and LLM agent flexibility.