OccuBench is a new benchmark for AI agents on real-world occupational tasks via LLM-driven simulators, showing no model dominates all industries, implicit faults are hardest, and larger models with more reasoning perform better.
hub Canonical reference
Cl-bench: A benchmark for context learning
Canonical reference. 80% of citing Pith papers cite this work as background.
20
Pith papers citing it
Background
80% of classified citations
hub tools
citation-role summary
background 5
citation-polarity summary
years
2026 20roles
background 5representative citing papers
Archer automates agentic code review for LLVM optimizations and reports finding semantic bugs in 21% of recent open PRs and 11% of closed PRs.
SafePyramid is a three-level benchmark showing frontier LLMs identify all violated rules in only 54.0%, 35.3%, and 12.9% of cases on L0, L1, and L2 respectively, indicating in-context policy guardrailing remains difficult.
Introduces VG-GUIBench benchmark and TASKER keyframe extraction algorithm that improves performance on VideoQA and video-guided agentic tasks.
LongMemEval-V2 is a new benchmark where AgentRunbook-C reaches 72.5% accuracy on long-term agent memory tasks, beating RAG baselines at 48.5% and basic coding agents at 69.3%.
EVPO adaptively switches between critic-based and batch-mean advantage estimation using batch-level explained variance to provably achieve no greater variance than the better of PPO or GRPO at every step.
DEONTICBENCH is a new benchmark of 6,232 deontic reasoning tasks from U.S. legal domains where frontier LLMs reach only ~45% accuracy and symbolic Prolog assistance plus RL training still fail to solve tasks reliably.
Benchmark Agent is an autonomous agentic system that constructs benchmarks for LLMs and MLLMs via query analysis, subtask design, annotation and quality control, yielding 15 benchmarks with minimal human input.
Introduces BeliefTrack benchmark diagnosing three CBM failures in LLMs and shows RL with belief-state rewards cuts failure rates by 70.9% while representation steering cuts them by 46.1%.
PEEK maintains a constant-sized context map via a programmable cache policy to give LLM agents persistent orientation knowledge about recurring external contexts, yielding 6-34% gains and lower cost than prior prompt-learning methods.
EvoMemBench evaluates 15 memory methods for LLM agents and finds long-context baselines competitive with no single memory approach working consistently across settings.
MMCL-Bench shows that even the strongest frontier multimodal models solve fewer than one-third of tasks requiring recovery and application of visual rules, procedures, and empirical patterns.
Introduces Personal VCL formalization and benchmark revealing LMM context gaps, plus an Agentic Context Bank baseline that boosts personalized visual reasoning.
Ctx2Skill automatically produces natural-language skill files via self-play between a challenger, a reasoner, and a judge, improving LLM performance on context-learning tasks.
CL-bench Life shows frontier language models achieve only 13.8% average success on real-life context tasks, with the best model at 19.3%.
SWE-MeM introduces adaptive memory management for coding agents via synthesized trajectories and Memory-aware GRPO, reporting 43.4% and 60.2% resolve rates on SWE-Bench Verified for 4B and 30B models while beating baselines on performance and token use.
Evo-MedAgent adds three evolving memory stores to LLM agents for chest X-ray diagnosis, raising MCQ accuracy from 0.68 to 0.79 on GPT-5-mini and 0.76 to 0.87 on Gemini-3 Flash without any training.
LLMs can execute specific technical steps in statistical proofs when given precise guidance but become unreliable for open-ended problem formulation or multi-step reasoning, relocating rather than reducing the demand for human expertise.
Proposes Knowledge Objects to externalize implicit AI knowledge for human verification, addressing a gap in current reliability methods.
Seed2.0 model series reports gains in reasoning, visual understanding, search, and reliability on intricate long-horizon tasks via an internal evaluation system.
citing papers explorer
-
OccuBench: Evaluating AI Agents on Real-World Professional Tasks via Language Environment Simulation
OccuBench is a new benchmark for AI agents on real-world occupational tasks via LLM-driven simulators, showing no model dominates all industries, implicit faults are hardest, and larger models with more reasoning perform better.
-
Archer: Towards Agentic Review for Compiler Optimizations
Archer automates agentic code review for LLVM optimizations and reports finding semantic bugs in 21% of recent open PRs and 11% of closed PRs.
-
SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing
SafePyramid is a three-level benchmark showing frontier LLMs identify all violated rules in only 54.0%, 35.3%, and 12.9% of cases on L0, L1, and L2 respectively, indicating in-context policy guardrailing remains difficult.
-
Bridging VideoQA and Video-Guided Agentic Tasks via Generalized Keyframe Extraction
Introduces VG-GUIBench benchmark and TASKER keyframe extraction algorithm that improves performance on VideoQA and video-guided agentic tasks.
-
LongMemEval-V2: Evaluating Long-Term Agent Memory Toward Experienced Colleagues
LongMemEval-V2 is a new benchmark where AgentRunbook-C reaches 72.5% accuracy on long-term agent memory tasks, beating RAG baselines at 48.5% and basic coding agents at 69.3%.
-
EVPO: Explained Variance Policy Optimization for Adaptive Critic Utilization in LLM Post-Training
EVPO adaptively switches between critic-based and batch-mean advantage estimation using batch-level explained variance to provably achieve no greater variance than the better of PPO or GRPO at every step.
-
DeonticBench: A Benchmark for Reasoning over Rules
DEONTICBENCH is a new benchmark of 6,232 deontic reasoning tasks from U.S. legal domains where frontier LLMs reach only ~45% accuracy and symbolic Prolog assistance plus RL training still fail to solve tasks reliably.
-
Benchmark Everything Everywhere All at Once
Benchmark Agent is an autonomous agentic system that constructs benchmarks for LLMs and MLLMs via query analysis, subtask design, annotation and quality control, yielding 15 benchmarks with minimal human input.
-
When Should Models Change Their Minds? Contextual Belief Management in Large Language Models
Introduces BeliefTrack benchmark diagnosing three CBM failures in LLMs and shows RL with belief-state rewards cuts failure rates by 70.9% while representation steering cuts them by 46.1%.
-
PEEK: Context Map as an Orientation Cache for Long-Context LLM Agents
PEEK maintains a constant-sized context map via a programmable cache policy to give LLM agents persistent orientation knowledge about recurring external contexts, yielding 6-34% gains and lower cost than prior prompt-learning methods.
-
EvoMemBench: Benchmarking Agent Memory from a Self-Evolving Perspective
EvoMemBench evaluates 15 memory methods for LLM agents and finds long-context baselines competitive with no single memory approach working consistently across settings.
-
MMCL-Bench: Multimodal Context Learning from Visual Rules, Procedures, and Evidence
MMCL-Bench shows that even the strongest frontier multimodal models solve fewer than one-third of tasks requiring recovery and application of visual rules, procedures, and empirical patterns.
-
Personal Visual Context Learning in Large Multimodal Models
Introduces Personal VCL formalization and benchmark revealing LMM context gaps, plus an Agentic Context Bank baseline that boosts personalized visual reasoning.
-
From Context to Skills: Can Language Models Learn from Context Skillfully?
Ctx2Skill automatically produces natural-language skill files via self-play between a challenger, a reasoner, and a judge, improving LLM performance on context-learning tasks.
-
CL-bench Life: Can Language Models Learn from Real-Life Context?
CL-bench Life shows frontier language models achieve only 13.8% average success on real-life context tasks, with the best model at 19.3%.
-
SWE-MeM: Learning Adaptive Memory Management for Long-Horizon Coding Agents
SWE-MeM introduces adaptive memory management for coding agents via synthesized trajectories and Memory-aware GRPO, reporting 43.4% and 60.2% resolve rates on SWE-Bench Verified for 4B and 30B models while beating baselines on performance and token use.
-
Evo-MedAgent: Beyond One-Shot Diagnosis with Agents That Remember, Reflect, and Improve
Evo-MedAgent adds three evolving memory stores to LLM agents for chest X-ray diagnosis, raising MCQ accuracy from 0.68 to 0.79 on GPT-5-mini and 0.76 to 0.87 on Gemini-3 Flash without any training.
-
Statistical Proof as a Window into Human-AI Collaboration: Practical Insights and a Community Agenda
LLMs can execute specific technical steps in statistical proofs when given precise guidance but become unreliable for open-ended problem formulation or multi-step reasoning, relocating rather than reducing the demand for human expertise.
-
Reliable AI Needs to Externalize Implicit Knowledge: A Human-AI Collaboration Perspective
Proposes Knowledge Objects to externalize implicit AI knowledge for human verification, addressing a gap in current reliability methods.
-
Seed2.0 Model Card: Towards Intelligence Frontier for Real-World Complexity
Seed2.0 model series reports gains in reasoning, visual understanding, search, and reliability on intricate long-horizon tasks via an internal evaluation system.