concept
active
concept:llavaLLaVA
Multimodal model demonstrating that projecting visual features into LLM with 2-layer MLP achieves state-of-the-art results
Neighborhood — ranked by edge-count
Papers (1)
paper
- The Platonic Representation Hypothesiscitesmentions
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Language model family used in cross-modal alignment experiments across multiple sizes
- Meta's open large language model cited as an example of the class of models under discussion
- Google's conversational AI, discussed as a system for generating chatbots, highlighting simulator nature.
- Most closely related work seeking to unify method overloading for generic functions dynamically using minimal operators; treats type as derived runtime property rather than formal compile-time abstraction.
- One of four LLMs selected for representation analysis; embedding dimension D=4096; used as demonstration model in scatter plots.
- A smaller recessed space within a room that forms a strong center.
- T5 model fine-tuned on Stanford Alpaca data; used to initialize Multimodal-CoT model weights.