CS PhD Student @ICatGT | Undergrad @UofIllinois
Atlanta, GA
Joined October 2018
- Tweets70
- Following395
- Followers165
- Likes725
Pinned Tweet
Real-world LLM interactions are conversational, yet LLMs are far more vulnerable to multi-turn than single-turn attacks.
Introducing DialTree: RL + tree search for automated multi-turn attack discovery.
📊 81.5% avg ASR@1 across 12 LLMs
📈 +44.2% over prior SOTA
#ICLR2026 [1/6]
Ruohao Guo retweeted
Updated CS 8803 "Large Language Model" class at @GeorgiaTech this year for 2026!
The reading list spans pretraining, MoE, reasoning, RL & self-play, agents, long-context, test-time scaling, diffusion LMs, safety, interpretability, and more.
👉 cocoxu.github.io/CS8803-LLM-…
🔗 docs.google.com/spreadsheets…
Ruohao Guo retweeted
Yesterday, my OpenAI collaborator and I gave a detailed talk on the Huggingface incident, our models creating "the message board", model misalignment, and more.
youtube.com/watch?v=87DyyMV0…
I hope it can answer a lot of the questions folks have, and we will release a full detailed postmortem at a later time!
Ruohao Guo retweeted
We are excited to announce Evaluation of Interactive Agents Workshop (IAEval) at NeurIPS 2026 in Atlanta.
As AI agents become more capable and engage with users on long, open-ended tasks, static benchmarks are no longer enough. We need evaluation methods that capture how agents behave in realistic, multi-turn interactions.
If your research is relevant, we invite you to submit. More details below.
🤖 Made with AI
Ruohao Guo retweeted
📢Excited to share a best-in-class end-to-end audio LLM that listen and speak with no loss in reasoning capability.
Building on Nemotron-Cascade-2, Audex tackles the core challenge that has bottlenecked audio model intelligence.
Check more details below!
🚀 Introducing Audex, a unified audio-text LLM for text, speech, sound, and music 🚀
🏆 Audex delivers best-in-class performance among open models across:
🎧 Audio understanding
🗣️ Speech recognition and translation
🔊 Text-to-speech
🎵 General audio generation
🔄 Speech-to-speech generation
🥇 Audex also achieves best-in-class results in math and code reasoning, alignment, and instruction following, even outperforming the text-only Qwen3.5-35B-A3B.
🧩 Minimalist architecture: a single 30B-A3B MoE model. Audio inputs are projected into the text embedding space, while text tokens and quantized audio output tokens are treated uniformly during generation.
🧠 Strong text intelligence, preserved
• Built on the Nemotron-Cascade-2-30B-A3B text backbone
• Trained with multi-stage SFT on blended audio-text data, Cascade RL, and multi-domain on-policy distillation
• The result: broad and SOTA audio capabilities with no regression in text intelligence.
🤗 Model:
👉 huggingface.co/nvidia/Nemotr…
📄 Technical report:
👉 arxiv.org/pdf/2607.05196
Ruohao Guo retweeted
“AI agents will outperform humans at almost all jobs by 2026–2027.” - The forecast is everywhere.
So we built the exam to test that claim, on real labor-market aligned work. On the hardest tier, top agents pass 2.6%.
Meet Agents' Last Exam (ALE), a rolling benchmark measuring whether agents can actually do real jobs. 🧵👇
Ruohao Guo retweeted
Knowledge is not evenly distributed across languages. For a given question, what if better information comes from another language?
🔥 We propose Language-Routed Policy Optimization (LRPO), an online reinforcement learning framework that lets LLMs explore and learn from cross-lingual generations during training.
#ICML2026
[5/6] In addition, DialTree shows highest attack efficiency using fewest queries. Its branch expansion effectively increases trajectory diversity. More analyses are discussed in the paper.
[6/6] Huge thanks to my amazing coauthors—Afshin Oroojlooy, Roshan Sridhar, @migballesteros, @alan_ritter, @DanRothNLP —for an incredible collaboration at @Oracle
@OracleCloud and @ICatGT.
Website: dialtree.github.io/
Paper: arxiv.org/abs/2510.02286
Ruohao Guo retweeted
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.
Ruohao Guo retweeted
I’m so tired of writing rebuttals to this kind of “lack of novelty” review: “This paper trivially combines A, B, and C, so the algorithmic novelty is limited.”
Technically, most (if not all) robotics papers are convex combinations of existing ideas.
I still deeply appreciate A+B+C papers—especially when they deliver:
- New capabilities: the “trivial combination” unlocks behaviors we simply couldn’t achieve before
- Sensible & organic design: A+B+C is clearly the right composition—not some arbitrary A′+B+C′
- Nontrivial interactions: careful analysis of the dynamics, coupling, or failure modes between A, B, C
- Rehabilitating old ideas: A was dismissed for years, but paired with modern B/C, it suddenly works—and teaches us why
- System-level & "interface" insight: the contribution is not any single piece, but how the pieces talk to each other
- Scaling laws or regimes: identifying when/why A+B+C works (and when it doesn’t)
- Engineering clarity: making something actually work robustly in the real world is not “trivial”
- New problem formulations: sometimes the real novelty is in the reformulation—only under this view does A+B+C make sense.
Maybe worth keeping these in mind when reviewing the next A+B+C paper : )
Ruohao Guo retweeted
We released Nemotron Cascade 2 30B A3B.
What makes this release especially meaningful to me is that it reflects a 1.5-year journey at NVIDIA around one core idea: improving AI math reasoning through self-improvement at test time.
Each project tackled a different part of that problem.
With AceMath (24 Q4), we built an external verifier model to identify the right solution during test-time scaling.
With AceReason (24 Q1-2), we scaled the reasoning capabilities of the model through RL so the model could spend more time reflecting while solving problems. Along the way, we found a general, simple and effective RL recipe that we’ve kept using since.
And now with Cascade 2 (25 Q1), we’ve pushed that effort further: the model can generate hypotheses, verify them, and refine them on its own. That self-improvement loop is what enabled IMO gold-level performance at 30B level.
From MATH500, to AIME, and now IMO Proof.
This team is THE BEST.
Technical report:
huggingface.co/papers/2603.1…
🚀 Introducing Nemotron-Cascade 2 🚀
Just 3 months after Nemotron-Cascade 1, we’re releasing Nemotron-Cascade 2: an open 30B MoE with 3B active parameters, delivering best-in-class reasoning and strong agentic capabilities.
🥇 Gold Medal-level performance on IMO 2025, IOI 2025, and ICPC World Finals 2025:
• Capabilities once thought achievable only by frontier proprietary models (e.g. Gemini Deep Think) or frontier-scale open models (i.e. DeepSeek-V3.2-Speciale-671B-A37B).
• Remarkably high intelligence density with 20× fewer parameters.
🏆 Best-in-class across math, code reasoning, alignment, and instruction following:
• Outperforms the latest Qwen3.5-35B-A3B (2026-02-24) and even larger Qwen3.5-122B-A10B (2026-03-11).
🧠 Powered by Cascade RL + multi-domain on-policy distillation:
• Significantly expand Cascade RL across a much broader range of reasoning and agentic domains than Nemotron-Cascade 1, while distilling from the strongest intermediate teacher models throughout training to recover regressions and sustain gains.
🤗 Model + SFT + RL data:
👉 huggingface.co/collections/n…
📄 Technical report:
👉 research.nvidia.com/labs/nem…
Ruohao Guo retweeted
🥈 Silver Medal at IOI 2025 & Outperforms DeepSeek-R1-0528 on LiveCodeBench.
Instead of mixing different tasks together, we scale *Cascade RL* to develop general LLMs in curriculum (RLFH -> Instruct -> Math -> Code -> SWE).
So many learnings, check out our report!👇
🚀 Introducing Nemotron-Cascade! 🚀
We’re thrilled to release Nemotron-Cascade, a family of general-purpose reasoning models trained with cascaded, domain-wise reinforcement learning (Cascade RL), delivering best-in-class performance across a wide range of benchmarks.
💻 Coding powerhouse
After RL, our 14B model:
• Surpasses DeepSeek-R1-0528 (671B) on LiveCodeBench v5/v6/Pro.
• Achieves silver-medal performance at IOI 2025 🥈.
• Reaches a 43.1% pass@1 on SWE-Bench Verified, and 53.8% with test-time scaling.
🧠 What is Cascade RL?
Instead of mixing heterogeneous prompts across domains, Cascade RL trains sequentially, domain by domain, which reduces engineering complexity, mitigates heterogeneous verification latencies, and enables domain-specific curricula and tailored hyperparameter tuning.
✨ Key insight
Using RLHF for alignment as a pre-step dramatically boosts complex reasoning—far beyond preference optimization. Subsequent domain-wise RLVR stages rarely hurt the benchmark performance attained in earlier domains and may even improve it, as illustrated in the following figure.
🤗 Models & training data 🔥
👉 huggingface.co/collections/n…
📄 Technical report with detailed training and data recipes
👉 arxiv.org/pdf/2512.13607
Ruohao Guo retweeted
At #NeurIPS2025 through Sunday. Come say hi and check out our posters on:
🔒Probabilistic reasoning for text anonymity estimation: Wednesday @ 11am
🤖 Efficient, self-improving agents: Friday @ 11am – at San Diego, CA
Ruohao Guo retweeted
Introducing early experience: using future states resulting from agent’s own action as scalable supervision to train itself - without reward🧠!
1️⃣Reward-free: can train directly in real-world environments.
2️⃣Better RL warm-start: when continued with RL, leads to higher final performance than imitation-only warm-ups.
3️⃣Data-efficient & scalable: outperforms imitation with even 1/8 data.👇
🌀Agent Learning via Early Experience🌀
📝: arxiv.org/abs/2510.08558
- SFT for agents is sparse; RL on long-horizons is hard
We provide new mid-training signals that work:
1) Implicit next state world modeling task
2) Self-reflection on alternate states
- Strong improvements over 8 environments and multiple model families
- Works well for subsequent RL!
🧵1/5
Ruohao Guo retweeted
Can LLM-simulated users replace expensive human evaluation for multi-turn conversations? Short answer: yes, if you model the user right.
With our SimulatorArena, we find that detailed user profiles (knowledge + message style) improve alignment with real human evaluation by 26% at <3% the cost.
#EMNLP2025 [1/6] 🧵
