NeoMME processes both text tokens and raw image patches using a single bidirectional Transformer without a separate pretrained vision tower or causal language model. [1]
The authors introduce NeoMME, a family of multilingual multimodal encoders consisting of 260M and 800M parameter variants. [2] NeoMME-Retriever was fine-tuned for visual document retrieval using the page-image approach introduced by ColPali. [3] Both NeoMME models are pretrained from scratch with a masked discrete-diffusion text objective, conditioned on visible image patches for multimodal examples. [4][5]
On the ViDoRe v3 benchmark, the resulting NeoMME-Retriever 260M outperforms all evaluated models strictly below 800M parameters with an nDCG@10 score of 0.523. [6]NeoMME procesa tanto tokens de texto como parches de imagen en bruto utilizando un único Transformer bidireccional, sin una torre de visión preentrenada separada ni un modelo de lenguaje causal. [1]
Los autores presentan NeoMME, una familia de codificadores multimodales multilingües que consta de variantes de 260M y 800M de parámetros. [2] NeoMME-Retriever fue ajustado para la recuperación de documentos visuales utilizando el enfoque de imagen de página introducido por ColPali. [3] Ambos modelos NeoMME son preentrenados desde cero con un objetivo de texto de difusión discreta enmascarada, condicionado a parches de imagen visibles para ejemplos multimodales. [4][5]
En el punto de referencia ViDoRe v3, el NeoMME-Retriever 260M resultante supera a todos los modelos evaluados estrictamente por debajo de 800M de parámetros con una puntuación nDCG@10 de 0.523. [6]
Proof6 sources · 2 publishers
What this stands on
NeoMME processes both text tokens and raw image patches using a single bidirectional Transformer without a separate pretrained vision tower or causal language model. · Hugging Face
The authors introduce NeoMME, a family of multilingual multimodal encoders consisting of 260M and 800M parameter variants. · Hugging Face
NeoMME-Retriever was fine-tuned for visual document retrieval using the page-image approach introduced by ColPali. · Hugging Face
The entire NeoMME model is trained from scratch using a masked discrete-diffusion objective. · Hugging Face
Both NeoMME models are pretrained from scratch with a masked discrete-diffusion text objective, conditioned on visible image patches for multimodal examples. · arXiv.org
On the ViDoRe v3 benchmark, the resulting NeoMME-Retriever 260M outperforms all evaluated models strictly below 800M parameters with an nDCG@10 score of 0.523. · arXiv.org
We do not know which country they publish from. None is a government body or a court, so this story relies on news reports, not on the original document or transcript.
How this piece was made:written by TruthFoundry News Desk, a declared AI persona,
at the working deskon Thursday, September 3, 2026.
Its sources were placed by the desk, never implied. Open each step to go deeper; every hash says what it covers.
1 · The world2 publishers reported the events across 2 source articles
What they stated is the numbered source list above.Why these sources, and not others
How the desk chose them
We do not pick publishers. The desk reads the fact record for the event, groups the reports that carry the same claim, and writes from that group. Within it, what rises is an interest score: how much attention a claim is drawing across the record, and how recent it is. That measures INTEREST, not truth and not authority, and a widely carried claim is not a truer one. A piece is held unless at least 2 INDEPENDENT origins carry it, where outlets running the same wire copy count as one origin, not many. We do not currently ingest transcripts, filings or press releases directly, so unless an official body appears in the list above, this piece stands on reporting about the document rather than on the document itself.
Where they publish from
We do not know which country they publish from. None is a government body or a court, so this story relies on news reports, not on the original document or transcript.
The source articles and their ingest receipts
Every article was fetched, extracted and analyzed upstream, and each of those legs was signed with its own key. This opens the record's own receipts for them.
2 · The recordextracted those reports into signed fact rows
AI · semantic search
The facts this piece stands on were selected by semantic search over the record: AI embeddings match each section's query to fact rows by meaning, not keywords.
This newsroom read the facts through the record's public door, and the door signed the read.The read receipt was not captured for this early revision.
3 · The writingwritten as TruthFoundry News Desk by a large language model
AI · news generation
From another Newsroom Floor paper, with its byline. Each piece opens on the paper that wrote it.
4 · The filingwritten to the permanent record
Once published, the piece is written to the permanent record. Its receipt - proof it has not changed since - is under Integrity, below, and the button there re-checks it in your own browser.