architecture · first seen 16 Oct 2026
encoder-free multimodal model
Multimodal language models that feed raw audio, pixel, or spectral inputs directly into the transformer without modality-specific encoders.
Mentions per week
PapersCodeDiscussion
Earliest sources
- Encoder-Free Multimodal LLMs Learn Implicit EncodersPapers · 16 Oct 2026 · David Harwath, Desh Raj, Xavier Velez, Sandra Arcos-Holzinger, Willie Jeng, Kare…
- Unbundling Multimodal Vision EncodingPapers · 12 Nov 2026 · Ellis Brown, Chris Hoang, Shengyi Qian, Shang-Wen Li, Mike Lewis, Marjan Ghazvin…
Latest sources
- Unbundling Multimodal Vision EncodingPapers · 12 Nov 2026 · Ellis Brown, Chris Hoang, Shengyi Qian, Shang-Wen Li, Mike Lewis, Marjan Ghazvin…
- Encoder-Free Multimodal LLMs Learn Implicit EncodersPapers · 16 Oct 2026 · David Harwath, Desh Raj, Xavier Velez, Sandra Arcos-Holzinger, Willie Jeng, Kare…