AutoResearchBench is a new benchmark showing top AI agents achieve under 10% success on complex scientific literature discovery tasks that demand deep comprehension and open-ended search.
hub Mixed citations
Qwen3.5: Towards native multimodal agents, February 2026
Mixed citation behavior. Most common role is method (43%).
22
Pith papers citing it
Method
43% of classified citations
hub tools
citation-role summary
method 3
background 2
baseline 1
dataset 1
citation-polarity summary
years
2026 22representative citing papers
SpaceDG is the first large-scale benchmark dataset (~1M QA pairs) simulating nine visual degradations in 3DGS-rendered scenes to measure and improve spatial intelligence robustness in MLLMs.
LMM-Track4D formulates a trajectory-grounded dialogue task, releases Track4D-Bench with 526 samples, and proposes RTGE encoding, TRK state token, and OSK-RA decoder to elicit better 4D spatiotemporal reasoning in LMMs.
Edit-Compass and EditReward-Compass are new unified benchmarks for fine-grained image editing evaluation and realistic reward modeling in reinforcement learning optimization.
NaiAD is a new dataset and framework for LLM-native advertising that uses decoupled generation and calibrated scoring to identify four semantic strategies for balancing user and commercial utilities.
Joint Consistency casts test-time aggregation as Ising-type energy minimization with pairwise LLM-judge interactions, subsuming voting methods and outperforming baselines across reasoning tasks.
RiskWebWorld is the first realistic interactive benchmark for GUI agents in e-commerce risk management, revealing a large gap between generalist and specialized models plus RL gains.
Memory Grafting improves language-model benchmarks by grafting offline hidden-state memory from a larger model into a recipient model using n-gram lookups and lightweight adapters, outperforming MoE and vanilla Engram baselines at 0.92B and 2.8B scales.
ClinSeekAgent automates active multimodal evidence seeking for clinical reasoning, improving LLM performance on raw EHR and CXR tasks while enabling distillation into smaller models.
OpenJarvis decomposes personal AI into Intelligence, Engine, Agents, Tools & Memory, and Learning primitives and applies LLM-guided spec search to produce on-device configurations that reach within 3.2 pp of cloud baselines on average across eight tasks.
PCAP conditions adversarial searches on multiple attacker personas to discover more diverse and transferable jailbreaks, yielding richer safety fine-tuning datasets that boost model robustness on GPT-OSS 120B.
TrajPrism introduces a multi-task benchmark with 300K real-world urban trajectories and 2.1M language-grounded task instances across three cities, plus proof-of-concept models showing large gaps versus geometry-only baselines.
EmbodiSkill uses skill-aware reflection on execution trajectories to update skills in embodied agents, achieving 93.28% success on ALFWorld with a frozen Qwen3.5-27B model, outperforming direct GPT-5.2 use by 31.58%.
Pre-trained MoE models exhibit up to 90% intra-expert activation sparsity that enables up to 2.5x faster MoE layer execution when exploited in the vLLM inference system.
LLM agents struggle to detect and act on implicit memory conflicts, with top models scoring 55.2% on the new STALE benchmark of 400 scenarios; CUPMem prototype strengthens state-aware revision.
DynT2I-Eval creates fresh prompts via dimension decomposition and dynamic sampling to evaluate text-to-image models on text alignment, quality, and aesthetics while maintaining a stable leaderboard.
PhyWorld improves temporal consistency and physical plausibility in video world models via flow matching fine-tuning followed by DPO on physics preference pairs, with reported gains on VBench and a custom physical-faithfulness benchmark.
SafeLens presents a fast-and-slow video guardrail framework that filters the SafeWatch dataset to 2.4% and adds Chain-of-Thought traces to achieve state-of-the-art moderation performance at reduced inference cost.
HTAF is a sigmoid-tanh composite that approximates the Heaviside function to allow stable gradient training of binary activation networks, yielding ICBMs with stable discretization and competitive performance on image tasks.
Squeeze Evolve is a multi-model orchestration framework that improves efficiency and performance in verifier-free evolutionary inference, cutting costs up to 3x while matching verifier-based methods on several benchmarks.
JoyAI-LLM Flash delivers a 48B MoE LLM with 2.7B active parameters per token via FiberPO RL and dense multi-token prediction, released with checkpoints on Hugging Face.
EXAONE 4.5 is a new open-weight multimodal model that matches general benchmarks and outperforms similar-scale models on document understanding and Korean contextual reasoning.
citing papers explorer
-
Persona-Conditioned Adversarial Prompting: Multi-Identity Red-Teaming for Adversarial Discovery and Mitigation
PCAP conditions adversarial searches on multiple attacker personas to discover more diverse and transferable jailbreaks, yielding richer safety fine-tuning datasets that boost model robustness on GPT-OSS 120B.
-
TrajPrism: A Multi-Task Benchmark for Language-Grounded Urban Trajectory Understanding
TrajPrism introduces a multi-task benchmark with 300K real-world urban trajectories and 2.1M language-grounded task instances across three cities, plus proof-of-concept models showing large gaps versus geometry-only baselines.
-
EmbodiSkill: Skill-Aware Reflection for Self-Evolving Embodied Agents
EmbodiSkill uses skill-aware reflection on execution trajectories to update skills in embodied agents, achieving 93.28% success on ALFWorld with a frozen Qwen3.5-27B model, outperforming direct GPT-5.2 use by 31.58%.