5arXiv cs.CL (Computation and Language)·
Analogical Deep Research: LLM agents for foresight via historical analogy retrieval
Researchers introduce Analogical Deep Research (ADR), a new task and benchmark (ADR-bench) for evaluating whether LLM agents can find and leverage structural historical analogies for foresight analysis. The paper identifies a core failure mode: agents match on surface features rather than causal mechanisms. The authors propose CANA (Causal Analogical Researcher), an agentic framework using structural decomposition and reflective feedback, achieving up to 10% improvement in analogy generation and outperforming state-of-the-art deep research agents on ADR-bench.
Related guides (2)
Related events (8)
AARRI-Bench evaluates frontier LLMs and agents on granular research-intern-level tasks
Researchers introduce AARR (Act As a Real Researcher), a new benchmark series targeting whether AI agents can emulate the professionalism, thoroughness, and nuanced judgment of human researchers in granular research scenarios—not just macro-level task execution. The first benchmark, AARRI-Bench, tests frontier models and agentic harnesses, finding that even the best configuration (Mini-SWE-Agent with Claude Opus 4.7) achieves only 68.3% success, frequently missing subtle but critical details obvious to human researchers. The work argues that closing the gap requires deeper modeling of research behavior rather than more complex scaffolding.
Study compares human and LLM active causal reasoning, finding LLMs less efficient but near human-level on conjunctive rules
A new arXiv paper investigates whether active exploration reduces the 'conjunctive handicap' in causal learning, using a blicket detector task with adult participants who could freely intervene to identify causal objects. Results show active exploration substantially improves human conjunctive causal reasoning, though conjunctive rules still require more tests than disjunctive ones. State-of-the-art LLMs approach human-level hypothesis inference accuracy but show less efficient exploration strategies and similar conjunctive-disjunctive performance gaps, raising questions about the nature of LLM causal reasoning.
CausaLab: Scalable Benchmark for Interactive Causal Discovery by LLM Agents
CausaLab is a new evaluation environment that tests LLM agents on interactive causal discovery tasks, requiring them to recover both causal graphs and structural equations from synthetic laboratory episodes governed by randomly sampled structural causal models (SCMs). The benchmark separates predictive accuracy from genuine causal understanding, revealing a persistent gap: GPT-5.2-high achieves 92% task accuracy in a 6-node observational setting but only 0.471 all-edge F1 for mechanism recovery. Mixed observation-intervention strategies improve structural fidelity, while pure intervention strategies underperform on both metrics. Premature stopping is identified as a key agent weakness, partially mitigated by prompting models to verify hypothesis-data consistency.
Benchmark and study of LLM agents in deliberative multi-agent collaboration under partial observability
Researchers introduce a scalable benchmark for evaluating LLM agents on cooperative joint decision-making tasks where agents must exchange information under partial and asymmetric observations to reach a shared decision. A systematic evaluation of representative LLMs finds that state-of-the-art models still struggle with complex deliberative collaboration, failing in either information alignment or downstream reasoning even with external mathematical tools. Diagnostic analysis also reveals that deliberation can enable reflection and error correction, sometimes outperforming centralized baselines, offering a nuanced picture of multi-agent LLM capabilities.
LLawCo framework teaches embodied multi-agent LLMs to derive and follow cooperation laws
Researchers from MERL propose LLawCo (Learning Laws of Cooperation), a framework that enables embodied LLM-based agents to autonomously align with partners and task objectives in decentralized, partially observable environments. Agents reflect on past failures to extract misaligned behavioral patterns and derive high-level behavioral laws (e.g., 'Talk when necessary', 'Wait for partner'), which are incorporated into reasoning via supervised fine-tuning. The authors also introduce PARTNR-Dialog, a new large-scale multi-agent communicative planning benchmark, and report average success rate improvements of 4.5% on PARTNR-Dialog and 6.8% on TDW-MAT over state-of-the-art open-source communicative agent frameworks across four backbone LLMs.
RA-RFT: Retrieval-Augmented Reinforcement Fine-Tuning teaches LLMs to reason by analogy
Researchers propose Retrieval-Augmented Reinforcement Fine-Tuning (RA-RFT), a post-training framework that trains a retriever to rank contexts by expected reasoning benefit rather than semantic similarity, then fine-tunes a policy model via reinforcement learning using retrieved analogous demonstrations. The key insight is that reasoning-relevant retrieval surfaces complementary solution strategies rather than superficially similar problems. On mathematical reasoning benchmarks, RA-RFT improves AIME 2025 average@32 accuracy by 7.1 and 2.8 points over GRPO for Qwen3-1.7B and Qwen3-4B respectively, suggesting reasoning-aware retrieval is orthogonal to reward design and training curriculum improvements.
Action research documents 'Index Sickness' failure pattern in long-horizon LLM collaboration and proposes fix
A practitioner-researcher documents a failure mode called 'Index Sickness' observed across 391 consecutive LLM collaboration sessions on a real software project (Bang-v3): when symbolic identifier systems and rule-based System Prompts exceed a complexity threshold, LLMs abandon semantic grounding and produce internally consistent but reality-disconnected outputs. The paper names the underlying principle the 'Pang Principle (Semantic Vitality Law),' asserting that natural language with explicit purpose conveys higher information quality than symbolic expression. A proposed engineering fix, 'Baseline-Log Physical Separation,' reduced AI instruction volume by ~75% and eliminated recurrence over ~150 subsequent sessions. The work is action research rather than controlled experiment, but offers rare longitudinal empirical data on LLM degradation in long-horizon agentic workflows.
HypoArena benchmark evaluates LLMs on open-ended hypothesis discovery before conclusions are known
Researchers introduce Prospective Hypothesis Discovery (PHD), a new capability framing that tests whether LLMs can autonomously generate grounded, testable hypotheses from inconclusive evidence rather than answering pre-specified questions. The benchmark HypoArena includes 988 cases across six scientific and analytical domains, constructed via a novel Retrospective Context Regression pipeline that strips conclusions from expert documents. Evaluation of 15 frontier LLMs using a Bradley-Terry-Davidson arena framework reveals clear capability stratification and model-dependent effects of structured analytical prompting, with some models regressing under structured guidance. The work positions hypothesis formulation as a distinct, underexplored evaluation target for LLMs.