LLMEval-Logic is a solver-verified Chinese logical reasoning benchmark with 246 base and 190 hard items that shows frontier LLMs reach only 37.5% hard-item accuracy and 60.16% joint formalization score.
Title resolution pending
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
citation-role summary
background 1
dataset 1
citation-polarity summary
representative citing papers
LSP adds hierarchical hyperpriors over global sparsity and weight concentration parameters so that spike-and-slab models can discount inaccurate LLM weights while retaining gains when the weights are good.
SCALE-LoRA proposes a post-retrieval audit framework using sparse residual composition and disagreement-based reliability signals to improve open-pool LoRA adapter reuse on tasks like BIG-Bench Hard.
LightThinker++ adds explicit adaptive memory management and a trajectory synthesis pipeline to LLM reasoning, cutting peak token use by ~70% while gaining accuracy in standard and long-horizon agent tasks.
GRC unifies generation, retrieval, and compression in LLMs via meta latent tokens for single-pass execution with modular flexibility.
LLMs fail at extended counting of repeated characters due to finite internal states, with abrupt errors persisting across model scales and inference methods.
MiMo-V2-Flash is a 309B/15B MoE model trained on 27T tokens with hybrid attention and multi-teacher on-policy distillation that matches larger models like DeepSeek-V3.2 while enabling 2.6x faster decoding via repurposed MTP layers.
InternLM2 is a new open-source LLM that outperforms prior versions on 30 benchmarks and long-context tasks through scaled pre-training to 32k tokens and a conditional online RLHF alignment strategy.
citing papers explorer
-
LLMEval-Logic: A Solver-Verified Chinese Benchmark for Logical Reasoning of LLMs with Adversarial Hardening
LLMEval-Logic is a solver-verified Chinese logical reasoning benchmark with 246 base and 190 hard items that shows frontier LLMs reach only 37.5% hard-item accuracy and 60.16% joint formalization score.
-
LLM Sparsity Prior for Robust Feature Selection
LSP adds hierarchical hyperpriors over global sparsity and weight concentration parameters so that spike-and-slab models can discount inaccurate LLM weights while retaining gains when the weights are good.
-
SCALE-LoRA: Auditing Post-Retrieval LoRA Composition with Residual Merging and View Reliability
SCALE-LoRA proposes a post-retrieval audit framework using sparse residual composition and disagreement-based reliability signals to improve open-pool LoRA adapter reuse on tasks like BIG-Bench Hard.
-
LightThinker++: From Reasoning Compression to Memory Management
LightThinker++ adds explicit adaptive memory management and a trajectory synthesis pipeline to LLM reasoning, cutting peak token use by ~70% while gaining accuracy in standard and long-horizon agent tasks.
-
GRC: Unifying Reasoning-Driven Generation, Retrieval and Compression
GRC unifies generation, retrieval, and compression in LLMs via meta latent tokens for single-pass execution with modular flexibility.
-
Language models fail at extended rule following
LLMs fail at extended counting of repeated characters due to finite internal states, with abrupt errors persisting across model scales and inference methods.
-
MiMo-V2-Flash Technical Report
MiMo-V2-Flash is a 309B/15B MoE model trained on 27T tokens with hybrid attention and multi-teacher on-policy distillation that matches larger models like DeepSeek-V3.2 while enabling 2.6x faster decoding via repurposed MTP layers.
-
InternLM2 Technical Report
InternLM2 is a new open-source LLM that outperforms prior versions on 30 benchmarks and long-context tasks through scaled pre-training to 32k tokens and a conditional online RLHF alignment strategy.