Daniel Cer @daniel_m_cer

Research Scientist at @GoogleAI, @googIeresearch. scholar.google.com/citations?user… California, USA Joined March 2012

Tweets

178
Followers

563
Following

809
Likes

3K

Tony Wu @tonywu_71

3 weeks ago

Similarity maps also works for the Hf-native ColQwen2 model! 🤗 I have created a cookbook to quickly try this out: github.com/tonywu71/colpa…

Leonie @helloiamleonie

3 weeks ago

Similarity maps also works for the Hf-native ColQwen2 model! 🤗 I have created a cookbook to quickly try this out: github.com/tonywu71/colpa…

4 23 130 9K 69

Download Image

0 5 20 1K 9

Sigrid Jin | Jin Hyung Park 🪄 👧 @sigridjin_eth

a month ago

@daniel_m_cer Here is the Python Implementation. github.com/sigridjineth/c…

0 2 3 87 1

Our Gemini model just won a gold medal at the IMO 2025. It’s a massive milestone for AI, and I’m so proud to have played a part. My work focused on i) Post Training the core model that was used in the IMO effort and ii) inference-time scaling, which was a significant factor in…

Google DeepMind @GoogleDeepMind

2 months ago

155 773 4K 1.1M 704

Download Image

2 9 66 8K 6

Omar Khattab @lateinteraction

2 months ago

📢 If you’re at #SIGIR2025 this week, make sure to be at Luca Scheerer’s paper talk: “WARP: An Efficient Engine for Multi-Vector Retrieval” (Wednesday 11am) WARP makes PLAID, the famous ludicrously fast ColBERT engine, another 3x faster on CPUs. With the usual ColBERT quality!

2 20 126 34K 69

Download Image

tomaarsen @tomaarsen

2 months ago

‼️Sentence Transformers v5.0 is out! The biggest update yet introduces Sparse Embedding models, encode methods improvements, Router module for asymmetric models & much more. Sparse + Dense = 🔥 hybrid search performance! Details in 🧵

6 65 480 52K 304

Download Image

Nandan Thakur @beirmug

2 months ago

We had a successful participation of over 45+ teams & 150+ runs last year in TREC RAG 2024! 🔥🔥 We are back with @TREC_RAG 2025 this year! Make sure you participate in one of the RAG tracks this year to help accelerate RAG evaluation! ⚡⚡ Signup here: docs.google.com/forms/d/e/1FAI…

TREC RAG @ 2025 @TREC_RAG

2 months ago

1 2 13 2K 4

Download Image

1 3 8 889 4

Harvey Yiyun Fu @harveyiyun

3 months ago

LLMs excel at finding surprising “needles” in very long documents, but can they detect when information is conspicuously missing? 🫥AbsenceBench🫥 shows that even SoTA LLMs struggle on this task, suggesting that LLMs have trouble perceiving “negative space” in documents. paper:…

11 32 159 26K 69

Download Image

Oscar Hong @oscrhong

2 months ago

Interesting tidbit from prof @chrmanning: The first mention of “Large Language Model” comes from a 1998 NLP workshop Taiwan! Paper by Chun-Liang Chen, Bo-Ren Bai, Lee-Feng Chien, Lin-Shan Lee. “Large” in 1998 = 20M word corpus

1 7 12 5K 9

Download Image

Google Research @GoogleResearch

2 months ago

Neural embedding models have become a cornerstone of modern information retrieval. Today we introduce MUVERA, a state-of-the-art retrieval algorithm that reduces complex multi-vector retrieval back to single-vector maximum inner product search. More →goo.gle/4k8YRlN

6 79 402 60K 212

Download Image

Sumit @_reachsumit

3 months ago

Video-ColBERT: Contextualized Late Interaction for Text-to-Video Retrieval Introduces a bi-encoder approach that performs fine-grained token-wise interaction at both spatial and temporal levels using modified MaxSim operations and dual sigmoid loss. 📝arxiv.org/abs/2503.19009

0 3 24 2K 15

David Wan @meetdavidwan

3 months ago

Excited to share our new work, CLaMR! 🚀 We tackle multimodal content retrieval by jointly considering video, speech, OCR, and metadata. CLaMR learns to dynamically pick the right modality for your query, boosting retrieval by 25 nDCG@10 over single modality retrieval! 🧐…

1 66 184 34K 123

Download Image

Ben Clavié @bclavie

3 months ago

Multimodal RAG: Just use ColPali/DSE then pass your screenshots to the LLM This is the dream, but how well do LLMs read text contained in images? We wanted to know, so we tried a simple thing: do results change on evals when using screenshots rather than text as input? Yes.

16 82 438 70K 409

Download Image

Raphaël Sourty @raphaelsrty

3 months ago

I'm thrilled to announce the release of FastPlaid ! 🚀🚀 FastPlaid is a high-performance engine for multi-vector search, built from the ground up in Rust (with the help of Torch C++)⚡️ You can view FastPlaid as the counterpart of Faiss for multi vectors.

10 38 249 39K 185

Tony Wu @tonywu_71

3 months ago

🚀 ColQwen2 just dropped in Transformers! 🤗 Say goodbye to brittle OCR pipelines — now you can retrieve documents directly in the visual space with just a few lines of code. Perfect for your visual RAG workflows. Smarter, simpler, faster. Let's dive in! 👇 (1/N 🧵)

9 94 579 47K 748

Download Image

David Bau @davidbau

3 months ago

Dear MAGA friends, I have been worrying about STEM in the US a lot, because right now the Senate is writing new laws that cut 75% of the STEM budget in the US. Sorry for the long post, but the issue is really important, and I want to share what I know about it. The entire…

25 71 472 122K 137

Sumit @_reachsumit

3 months ago

R3-RAG: Learning Step-by-Step Reasoning and Retrieval for LLMs via Reinforcement Learning Introduces RL to teach LLMs adaptive search intensity scaling, i.e. dynamically adjusting retrieval depth based on problem complexity. 📝arxiv.org/abs/2505.23794 👨🏽‍💻github.com/Yuan-Li-FNLP/R…

0 2 14 742 10

Manuel Faysse @ManuelFaysse

3 months ago

🚨 Context matters for effective retrieval—but most embedding models cannot leverage crucial information outside of the passage they embed. Our new paper "Context Is Gold to Find the Gold Passage" explores how context-aware embeddings can be trained to boost performance! 🧵(1/N)

7 30 174 20K 185

Download Image

Younggyo Seo @younggyoseo

3 months ago

Excited to present FastTD3: a simple, fast, and capable off-policy RL algorithm for humanoid control -- with an open-source code to run your own humanoid RL experiments in no time! Thread below 🧵

15 117 557 125K 299

Download Video

Nandan Thakur @beirmug

4 months ago

Did you know that fine-tuning retrievers & re-rankers on large but unclean training datasets can harm their performance? 😡 In our new preprint, we re-examine popular IR training data quality by pruning datasets and identifying and relabeling 𝐟𝐚𝐥𝐬𝐞-𝐧𝐞𝐠𝐚𝐭𝐢𝐯𝐞𝐬! 🏷️

2 19 85 10K 46

Download Image

Omar Khattab @lateinteraction

4 months ago

Google folks continues to do awesome late interaction work. Compared to vanilla ColBERT, a version of this new “CRISP achieves an 11x reduction in the number of vectors—with only a 3.6% quality loss”.