SimWorld Studio deploys an evolving coding agent to create adaptive 3D environments that co-evolve with embodied learners, delivering 18-point success-rate gains over fixed environments in navigation benchmarks.
hub Mixed citations
Qwen3.5: Towards native multimodal agents, February 2026
Mixed citation behavior. Most common role is background (40%).
21
Pith papers citing it
Background
40% of classified citations
hub tools
citation-role summary
background 4
method 3
dataset 2
baseline 1
citation-polarity summary
years
2026 21representative citing papers
Introduces APM benchmark with hidden randomized mappings to evaluate LLM style personalization methods fairly, finding routing most reliable while overall personalization remains challenging.
GPASV extends priority-aware Shapley values to arbitrary directed weighted graphs with penalization of violations and applies it to LLM ensemble valuation on the cyclic Chatbot Arena preference graph.
A genome-conditioned 4B LLM agent predicts microbial life boundaries and matches larger frontier models via token fusion, tool use, and a counterfactual gene-grounding reward.
Omnimodal LLMs encode premise-perception mismatches in hidden states yet almost never reject false textual claims, exposing a representation-action gap that is modality-asymmetric and prompt-resistant.
ProofRank benchmark shows substantial differences in LLM proof quality not captured by correctness, with trade-offs between quality metrics and accuracy.
CollabVR improves video reasoning performance by coupling vision-language models and video generation models in a closed-loop step-level collaboration that detects and repairs generation failures.
CUDAHercules benchmark demonstrates that leading LLMs generate functional CUDA code but fail to recover expert-level optimization strategies needed for peak performance on Ampere, Hopper, and Blackwell GPUs.
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.
TreeText-CTS builds source-traceable tree-path evidence from multi-scale EHR windows via frozen XGBoost, selects subsets, and uses an LM encoder to reach top AUPRC among text-based interfaces on mortality and sepsis tasks while staying competitive with numerical models.
Vision-OPD transfers an MLLM's privileged regional perception to its full-image policy through on-policy token-level self-distillation, yielding competitive results on fine-grained visual benchmarks.
MindVLA-U1 is the first unified streaming VLA architecture that surpasses human drivers on WOD-E2E planning metrics while matching VA latency and preserving language interfaces.
PrivacySIM shows that conditioning LLMs on user personas like demographics and attitudes improves simulation of privacy choices but reaches only 40.4% accuracy against real responses from 1,000 users.
Shepherd provides a reversible execution trace substrate for LLM agents that enables meta-agents to inspect and transform runs, yielding reported gains on coding and terminal benchmarks via supervision, counterfactual repair, and RL credit assignment.
DataMaster deploys an AI agent to autonomously engineer data via tree search over external sources, shared candidate pools, and memory of past outcomes, yielding 32% higher medal rates on MLE-Bench Lite and a small GPQA gain over the base instruct model.
Reformulating 53 visual reasoning tasks in polar coordinates causes frontier MLLMs to drop from 70-83% to 31-39% accuracy while preserving logical equivalence, revealing a Cartesian shortcut in current benchmarks.
Meow-Omni 1 is a quad-modal MLLM that fuses video, audio, physiological time-series, and text to achieve 71.16% accuracy on feline intent recognition in the new MeowBench benchmark.
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%.
Mix-Quant quantizes prefilling to NVFP4 and keeps BF16 for decoding in agentic LLMs, achieving up to 3x prefilling speedup while largely preserving task performance on long-context and agentic benchmarks.
TwiSTAR learns to switch between fast SID retrieval and slow rationale-generating reasoning in generative recommendation, yielding better accuracy-latency trade-offs on three datasets.
MathArena is broadened into a maintained platform with new benchmarks for proofs, research questions, and formal verification, where GPT-5.5 scores 98% on 2026 USAMO and 74% on research-level tasks.
citing papers explorer
-
SimWorld Studio: Automatic Environment Generation with Evolving Coding Agent for Embodied Agent Learning
SimWorld Studio deploys an evolving coding agent to create adaptive 3D environments that co-evolve with embodied learners, delivering 18-point success-rate gains over fixed environments in navigation benchmarks.
-
APM: Evaluating Style Personalization in LLMs with Arbitrary Preference Mappings
Introduces APM benchmark with hidden randomized mappings to evaluate LLM style personalization methods fairly, finding routing most reliable while overall personalization remains challenging.
-
Generalized Priority-Aware Shapley Value
GPASV extends priority-aware Shapley values to arbitrary directed weighted graphs with penalization of violations and applies it to LLM ensemble valuation on the cyclic Chatbot Arena preference graph.
-
GGBound: A Genome-Grounded Agent for Microbial Life-Boundary Prediction
A genome-conditioned 4B LLM agent predicts microbial life boundaries and matches larger frontier models via token fusion, tool use, and a counterfactual gene-grounding reward.
-
Senses Wide Shut: A Representation-Action Gap in Omnimodal LLMs
Omnimodal LLMs encode premise-perception mismatches in hidden states yet almost never reject false textual claims, exposing a representation-action gap that is modality-asymmetric and prompt-resistant.
-
Not All Proofs Are Equal: Evaluating LLM Proof Quality Beyond Correctness
ProofRank benchmark shows substantial differences in LLM proof quality not captured by correctness, with trade-offs between quality metrics and accuracy.
-
CollabVR: Collaborative Video Reasoning with Vision-Language and Video Generation Models
CollabVR improves video reasoning performance by coupling vision-language models and video generation models in a closed-loop step-level collaboration that detects and repairs generation failures.
-
CUDAHercules: Benchmarking Hardware-Aware Expert-level CUDA Optimization for LLMs
CUDAHercules benchmark demonstrates that leading LLMs generate functional CUDA code but fail to recover expert-level optimization strategies needed for peak performance on Ampere, Hopper, and Blackwell GPUs.
-
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.
-
TreeText-CTS: Compact, Source-Traceable Tree-Path Evidence for Irregular Clinical Time-Series Prediction
TreeText-CTS builds source-traceable tree-path evidence from multi-scale EHR windows via frozen XGBoost, selects subsets, and uses an LM encoder to reach top AUPRC among text-based interfaces on mortality and sepsis tasks while staying competitive with numerical models.
-
Vision-OPD: Learning to See Fine Details for Multimodal LLMs via On-Policy Self-Distillation
Vision-OPD transfers an MLLM's privileged regional perception to its full-image policy through on-policy token-level self-distillation, yielding competitive results on fine-grained visual benchmarks.
-
MindVLA-U1: VLA Beats VA with Unified Streaming Architecture for Autonomous Driving
MindVLA-U1 is the first unified streaming VLA architecture that surpasses human drivers on WOD-E2E planning metrics while matching VA latency and preserving language interfaces.
-
PrivacySIM: Evaluating LLM Simulation of User Privacy Behavior
PrivacySIM shows that conditioning LLMs on user personas like demographics and attitudes improves simulation of privacy choices but reaches only 40.4% accuracy against real responses from 1,000 users.
-
Shepherd: Enabling Programmable Meta-Agents via Reversible Agentic Execution Traces
Shepherd provides a reversible execution trace substrate for LLM agents that enables meta-agents to inspect and transform runs, yielding reported gains on coding and terminal benchmarks via supervision, counterfactual repair, and RL credit assignment.
-
DataMaster: Data-Centric Autonomous AI Research
DataMaster deploys an AI agent to autonomously engineer data via tree search over external sources, shared candidate pools, and memory of past outcomes, yielding 32% higher medal rates on MLE-Bench Lite and a small GPQA gain over the base instruct model.
-
The Cartesian Shortcut: Re-evaluate Vision Reasoning in Polar Coordinate Space
Reformulating 53 visual reasoning tasks in polar coordinates causes frontier MLLMs to drop from 70-83% to 31-39% accuracy while preserving logical equivalence, revealing a Cartesian shortcut in current benchmarks.
-
Meow-Omni 1: A Multimodal Large Language Model for Feline Ethology
Meow-Omni 1 is a quad-modal MLLM that fuses video, audio, physiological time-series, and text to achieve 71.16% accuracy on feline intent recognition in the new MeowBench benchmark.
-
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%.
-
Mix-Quant: Quantized Prefilling, Precise Decoding for Agentic LLMs
Mix-Quant quantizes prefilling to NVFP4 and keeps BF16 for decoding in agentic LLMs, achieving up to 3x prefilling speedup while largely preserving task performance on long-context and agentic benchmarks.
-
TwiSTAR:Think Fast, Think Slow, Then Act,Generative Recommendation with Adaptive Reasoning
TwiSTAR learns to switch between fast SID retrieval and slow rationale-generating reasoning in generative recommendation, yielding better accuracy-latency trade-offs on three datasets.
-
Beyond Benchmarks: MathArena as an Evaluation Platform for Mathematics with LLMs
MathArena is broadened into a maintained platform with new benchmarks for proofs, research questions, and formal verification, where GPT-5.5 scores 98% on 2026 USAMO and 74% on research-level tasks.