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.
Gpt-4o system card
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
citation-role summary
background 2
baseline 1
citation-polarity summary
representative citing papers
CoLVR uses latent contrastive objectives with angle-based perturbation and RL trajectory rewards to increase exploratory visual reasoning in MLLMs, delivering 5-8% gains on VSP, Jigsaw, and MMStar benchmarks.
LLM-based congestion control reduces latency by up to 50% with less than 0.3% throughput sacrifice versus traditional CCAs in static and dynamic network emulations.
RewardBench 2 is a new benchmark that supplies challenging fresh human prompts for reward model evaluation, yielding lower average scores but higher correlation with downstream best-of-N sampling and RLHF training performance.
OmniSelect is a training-free, modality-adaptive token pruning framework that dynamically selects Audio-Centric, Video-Centric, or Uniform compression regimes using AudioCLIP cross-modal relevance scores and then applies adaptive fine-grained pruning within temporal groups.
VideoLLaMA3 uses a vision-centric training paradigm and token-reduction design to reach competitive results on image and video benchmarks.
citing papers explorer
-
RewardBench 2: Advancing Reward Model Evaluation
RewardBench 2 is a new benchmark that supplies challenging fresh human prompts for reward model evaluation, yielding lower average scores but higher correlation with downstream best-of-N sampling and RLHF training performance.