Postdoc @ the Ohio State University
Joined July 2023
- Tweets69
- Following139
- Followers74
- Likes165
Every time reading reviews teaches me something new. Learned today that a new evaluation protocol should not be submitted to the evaluation and dataset track. @OpenAI @AnthropicAI when can your tools be trustworthy enough so reviewers really don’t need to check the correctness?
Jianyang Gu retweeted
📢 Reviewer Recruitment —CV4Ecology @ECCV 2026
Interested in contributing to the CV × Ecology community? Sign up here:
forms.gle/DPNboDfpjn45rS9b6
Workshop: cv4e-workshop.github.io/
Please help us spread the word! 🌿🐟📷
#CV4Ecology #ComputerVision #Ecology
Jianyang Gu retweeted
Our Archival Track deadline is coming up soon and has been extended to July 15th!
Looking forward to another great round of submissions 🙂
Jianyang Gu retweeted
🚀 Introducing our latest research: AgentCL: Toward Rigorous Evaluation of Continual Learning in Language Agents
Continual learning for language agents has not yet been clearly defined. How should we evaluate their ability to continually learn from experience and improve themselves on complex, long-horizon tasks?
- Traditional continual learning provides a useful perspective on the plasticity–stability trade-off, but its formulation does not naturally extend to non-parametric learning paradigms today.
- Many recent works on agent memory still focus on retrieval and reasoning over long (static) contexts, rather than on how agents can reuse experience from complex agentic tasks.
Across coding, deep research, and language understanding and reasoning tasks, we show that carefully designed task streams and metrics are essential for understanding continual learning in language agents.
📄 Paper: arxiv.org/abs/2606.02461
🤗 Dataset: huggingface.co/datasets/osun…
Jianyang Gu retweeted
We trained a ~frontier Deep Research Agent on academic budget
> 32 H100s
> 8K synthetic samples
> fully open training infra + recipe (SFT, mid-training, RL)
> models of diff sizes (2B -> 35B) ready to use out of the box
This is yet another demonstration of how the frontier of AI is changing. We have reached a point where open models + a small capable team + a few hundred Ks can produce specialized models with ~frontier capabilities. The future of AI doesn’t have to be held in a chokehold by a handful of closed models.
We've open-sourced everything we've built and learned from this project. Hope it helps the community build more!
📌 Project: osu-nlp-group.github.io/QUES…
📌 Paper: arxiv.org/abs/2605.24218
📌 Code: github.com/OSU-NLP-Group/QUE…
📌 Model Weights and Data: huggingface.co/collections/o…
📌 Demo: huggingface.co/spaces/osunlp…
Amazing effort led by @jianxie_ (our 1st year student!!), Tianhe Lin, Zilu Wang. joint with @hhsun1 and the @osunlp team. thanks @amazon Xiangjun Wang for a gift that covers the compute and fruitful discussion.
Jianyang Gu retweeted
🎉Excited to share that we have released the FishNet Challenge under the 3rd CV4Ecology Workshop @ ECCV 2026!
The challenge focuses on text-guided open-set fine-grained fish classification.🐠
Welcome to check it out and participate!
#CV4Ecology #ECCV26 #ComputerVision #Ecology
Jianyang Gu retweeted
Do VLMs actually understand 3D space 🌎?
Or are they exploiting shortcuts hidden in natural images?
🚀 Excited to share our new work:
Why Far Looks Up: Probing Spatial Representation in Vision-Language Models
@NVIDIAAI × @SeoulNatlUni × @OhioStateCSE
🧵👇
Submit to this year's CV4Ecology workshop at ECCV 2026!
We are excited to share that the CV4Ecology Workshop will return for its 3rd edition at #ECCV2026!
If you are working on the intersection of computer vision and ecology, we warmly welcome your submissions and participation.
Deadlines:
July 10, Archival
August 14, Non-Archival
Jianyang Gu retweeted
Gym environments have played a key role in advancing LMs and agents for general coding tasks. But how do we build them for scientific coding?
Introducing D3-Gym, the first automatically constructed dataset of verifiable environments for data-driven scientific discovery. 🧵
Always impressed by and fully agree with Yu’s vision on agents. Can’t wait for the product release and actually using it!
Introducing @NeoCognition, the agent lab for specialized intelligence.
Everyone needs experts, but human expertise does not scale.
Backed by $40M seed funding, we build self-learning agents that specialize across domains to make expertise abundant.
Jianyang Gu retweeted
Claude Mythos is suspected of being a Looped transformer (LT), but why are LT-based LLMs so powerful?
Our new finding: LT can perform implicit reasoning over their parametric knowledge, unlocking generalization to complex and unfamiliar questions compared to transformers ⤵️
Jianyang Gu retweeted
1/ Excited to share our #ICLR2026 paper on automatic image-level morphological trait annotation for organismal images.
Can we turn ecological images into grounded natural-language trait descriptions at scale?
Our answer: combine self-supervised vision features + sparse autoencoders + multimodal LLMs.
Jianyang Gu retweeted
Thanks @NVIDIAAI for featuring our work. Led by amazing students @vimar_gu @iamsamstevens at @osunlp
AI is helping scientists see nature in entirely new ways. 🔍
In collaboration with @OhioState, BioCLIP2 runs on NVIDIA accelerated computing to identify over a million species and reveal hidden patterns that support conservation and ecosystem health worldwide.
👉 nvda.ws/4v1RK5p
Jianyang Gu retweeted
AI is helping scientists see nature in entirely new ways. 🔍
In collaboration with @OhioState, BioCLIP2 runs on NVIDIA accelerated computing to identify over a million species and reveal hidden patterns that support conservation and ecosystem health worldwide.
👉 nvda.ws/4v1RK5p
Jianyang Gu retweeted
🚀Excited to share 𝗦𝗔𝗚𝗔!
Most AI for science asks: “How do we optimize better?”
We asked a different question: “How do we know we're optimizing for the right thing?”
Scientists don't arrive at perfect objectives — they discover them. SAGA automates exactly that: the messy, iterative process of figuring out what to optimize before how.
The design philosophy: a bi-level architecture that mirrors how scientists actually work:
🔁Outer loop: LLM agents analyze results, question current objectives, and evolve better ones
⚙️Inner loop: search hard under the objectives the outer loop proposes
SAGA is a generalist scientific discovery framework — the same system, applied across design of antibiotics, nanobodies, DNA sequences, inorganic materials, and chemical processes, with wet-lab validation🔬⚗️.
Check this out ⬇️
Jianyang Gu retweeted
Congrats to GPT-5.4 for achieving 92.8% success rate on Online-Mind2Web 🚀 I’m really impressed by its agentic capabilities.
I still remember when we released the benchmark about a year ago. Operator was around ~60% overall and only ~40% on complex tasks. Now, agents are getting close to near-perfect performance.
It's really a big step toward AGI. I'm curious when humans will finally be able to free their hands and let agents take over all the complex and tedious tasks. We’ll see, but I expect it sooner.
Jianyang Gu retweeted
Excited to see the first model with native computer-use capabilities from @OpenAI!
Glad to see multiple benchmarks done by @osunlp students (MMMU-Pro, SWE-Bench Pro, Online-Mind2Web) contributed to the evaluation.
Jianyang Gu retweeted
One static model does not fit all😭
We just dropped our latest work: Functional Neural Memory. Instead of static models, we generate custom "parameters" for every single input.
✅Prompt your model anytime
✅Instant personalization
✅Better instruction following
✅Flexible & dynamic memory (w/o memory bank✌️)
(🧵1/6)
