📊 We benchmarked top VLMs (GPT-4o, Claude, Gemini) vs. open-weight models (Qwen, LLaMA, Pixtral).

📈 Finetuned open models outperform proprietary ones:

Qwen2.5-VL-7B → FlowSim: 0.614
GPT-4o → FlowSim: 0.786
𝐐𝐰𝐞𝐧𝟐.𝟓-𝐕𝐋-𝟕𝐁 (𝐟𝐢𝐧𝐞𝐭𝐮𝐧𝐞𝐝) → 𝐅𝐥𝐨𝐰𝐒𝐢𝐦: 𝟎.𝟗𝟓𝟕

𝐖𝐡𝐲?

Workflow automation is powerful—but authoring flows is still complex, even with low-code tools.
💫𝐒𝐭𝐚𝐫𝐅𝐥𝐨𝐰 explores a simpler interface: 𝐣𝐮𝐬𝐭 𝐝𝐫𝐚𝐰 𝐢𝐭.

Imagine sketching a workflow on a whiteboard and getting a runnable flow in return.

🚀 New paper from our team at @servicenowresearch.bsky.social!⁣
⁣
💫𝐒𝐭𝐚𝐫𝐅𝐥𝐨𝐰: 𝐆𝐞𝐧𝐞𝐫𝐚𝐭𝐢𝐧𝐠 𝐒𝐭𝐫𝐮𝐜𝐭𝐮𝐫𝐞𝐝 𝐖𝐨𝐫𝐤𝐟𝐥𝐨𝐰 𝐎𝐮𝐭𝐩𝐮𝐭𝐬 𝐅𝐫𝐨𝐦 𝐒𝐤𝐞𝐭𝐜𝐡 𝐈𝐦𝐚𝐠𝐞𝐬⁣
We use VLMs to turn 𝘩𝘢𝘯𝘥-𝘥𝘳𝘢𝘸𝘯 𝘴𝘬𝘦𝘵𝘤𝘩𝘦𝘴 and diagrams into executable workflows 🖍️→⚙️⁣
⁣
🔗 arxiv.org/abs/2503.218...
📝 tinyurl.com/3utdbn97%E2%...
#Sketch2Flow #AI #VLM

📊 Our Results:

Multi-task instruction fine-tuning FTW! Our approach beats both BM25 and strong off-the-shelf encoder models across all retrieval tasks (in-distribution and out-of-distribution).

💡 The Challenge:

* RAG needs domain-specific knowledge
* Multiple apps = multiple retrievers = 💰
* Different types of data (steps, tables, fields, ...)

🚀 Excited to share our new work on making RAG actually work for enterprise applications!
We present a recipe to build a custom retriever that handles multiple retrieval tasks simultaneously for domain-specific RAG applications 🧵

Looking to build an LLM-powered app but finding it hard to make it robust? We’ve got you covered! Our new paper explores how Task Decomposition and Retrieval-Augmented Generation (RAG) can help you create reliable systems. 🧵👇