Saswat Das @ NeurIPS 2026 (ATL) (@WatIsDas) header
Saswat Das @ NeurIPS 2026 (ATL) (@WatIsDas) profile picture

Human. CS PhD Candidate at the RAISE Lab @CS_UVA. Responsible AI, Agentic AI, Privacy, Algorithmic Fairness, Security.

Charlottesville, VA
Joined July 2018
Photo by Saswat Das @ NeurIPS 2026 (ATL) (@WatIsDas)Photo by Saswat Das @ NeurIPS 2026 (ATL) (@WatIsDas)Photo by Saswat Das @ NeurIPS 2026 (ATL) (@WatIsDas)Photo by Saswat Das @ NeurIPS 2026 (ATL) (@WatIsDas)Photo by Saswat Das @ NeurIPS 2026 (ATL) (@WatIsDas)Photo by Saswat Das @ NeurIPS 2026 (ATL) (@WatIsDas)Photo by Saswat Das @ NeurIPS 2026 (ATL) (@WatIsDas)Photo by Saswat Das @ NeurIPS 2026 (ATL) (@WatIsDas)Photo by Saswat Das @ NeurIPS 2026 (ATL) (@WatIsDas)Photo by Saswat Das @ NeurIPS 2026 (ATL) (@WatIsDas)
Saswat Das @ NeurIPS 2026 (ATL) retweeted
We just made LLM inference up to 30% faster ⚡ Introducing DNeo, a new diffusion drafter designed in collaboration with @cerebras researchers. The secret? Your draft model needs a working memory, and the "right" context encoding mechanism. 📄 Paper: alphaxiv.org/pdf/2611.21925
9
40
21
211
23,997
Saswat Das @ NeurIPS 2026 (ATL) retweeted
📢New: SEABench Self-evolving LLM agents now improve after deployment by rewriting their own harness (controller instructions, memory procedures, tools and skills) in response to feedback. We ask whether a benign update that helps on one task can persist into later tasks where it no longer applies, and cause harm without any adversary involved. We call this endogenous misalignment. To study it, we built SEABench where an agent works as a personal assistant for a user, with access to their email, calendar, finances and health records. The agent handles everyday tasks, then it gets feedback. The agent is allowed to rewrite its own instructions, memory or tools. Then we freeze it and give it later tasks where the evolution no longer contextually fits. As a counterfactual, we run the same tasks on a copy that never updated itself. We use an automated search to find these cases and trajectories. The benchmark has 48 such sequences, testing for four kinds of failure such as: oversharing private data, carrying rules or information into the wrong context, hallucination and stating guesses as facts, and guardrail erosion by skipping approval steps. Across the discovered trajectories, due to locally useful updates, overall task completion indeed increased from 35.7% to 47.2%. But evolved agents also exhibited safety failures on 43.9% of tests, while the non-evolving baselines remained safe. We further propose a mitigation strategy that employs reasoning-trace monitoring. On held-out Kimi K2.5 traces, our monitor prevented 56 of 79 harmful outputs, with a 9.7% false-positive rate. Understanding safety and alignment implications of agentic behavior when they evolve over time is a timely research direction. SEABench aims to support this important discussion and further work on safer adaptation! Great collaboration with Saswat Das (@WatIsDas), Parvati Viswanathan, Daniel Donnelly, Chang Huang, and Ferdinando Fioretto (@nandofioretto)! Paper: arxiv.org/abs/2609.35596v2 Project page: seabench-endogenous-misalign…
12
9
2
50
3,374
Saswat Das @ NeurIPS 2026 (ATL) retweeted
Check out our new work on AI safety for self improving agents! We ask if self improving agents remain aligned to their original (safe) objective after self-improvement. Spoiler: they don’t! Interestingly, even if the improvement capability is purely benign, the agents diverge drastically from their original alignment objective. We provide an in depth analysis by allowing agents to self improve various aspects of their harness, in isolation, and release a new benchmark which we hope will be useful for the community to improve agentic safety. Paper: arxiv.org/pdf/2609.35596v2 Project page: seabench-endogenous-misalign… with Saswat Das (@WatIsDas), Parvati Viswanathan, Daniel Donnelly, Chang Huang, and Sahar Abdelnabi (@sahar_abdelnabi)
🤖 Made with AI
4
2
15
1,215
Saswat Das @ NeurIPS 2026 (ATL) retweeted
Incredible news and tremendous amount of work by @MasonNaka @abhinav_kumar26 and @WatIsDas . Colosseum is a good framework to build on top to study multi-agent behavior.
[#NeurIPS2026] I'm excited to share that our paper "Colosseum: Auditing Collusion in Cooperative Multi-Agent Systems" has been accepted to NeurIPS 2026 E&D Track 🎉 Our paper enables audits of multi-agent systems for collusion detection while identifying different types of collusion that may occur in practice. Our additional results show that when benign agents are given a secret communication channel these benign agents show propensities to collude. Thank you to my wonderful collaborators Abhinav Kumar (@abhinav_kumar26), Saswat Das (@WatIsDas), Sahar Abdelnabi (@sahar_abdelnabi), Saad Mahmud (@saadu_ai), Ferdinando Fioretto (@nandofioretto), Shlomo Zilberstein, and Eugene Bagdasarian (@ebagdasa)! 📰 [paper] arxiv.org/pdf/2602.15198 💻 [code] github.com/umass-ai-safety/c… See you all at NeurIPS, and I'd love to catch up with everyone working on multi-agent safety! #NeurIPS2026 #MultiAgentSystems #AI #Collusion #AISafety
3
1
14
1,861
Saswat Das @ NeurIPS 2026 (ATL) retweeted
Replying to @icmlconf
@icmlconf Review Process: Negligent reviewer whose concerns are already addressed in the paper/subjective+doesn't engage with the rebuttal. Two saying that their concerns are fully addressed but maintain their score. Possibly AI-generated PC metareview that ignores the rebuttal.
1
1
4
299
Saswat Das @ NeurIPS 2026 (ATL) retweeted
A few days ago, we shared our work showing that in multi-agent LLM systems, the biggest risk isn’t always one agent going rogue, it can be a whole group quietly coordinating on the wrong goal. 🚨 Now we’ve built a live demo so you can see that behavior in action. 👀⚔️ 🔗 Project Website : umass-ai-safety.github.io/co… 🔗 Interactive Demo : umass-ai-safety.github.io/co… Colosseum helps audit collusion in cooperative agent systems and detect: 🤝 direct collusion 🕵️ attempted collusion (agents coordinate in text, but actions don’t follow) 🎭 hidden collusion (collusive outcomes with no obvious signals) If you’re building agent teams, coordination risk should be a first-class safety concern. This work was done in collaboration with @MasonNaka, @WatIsDas, @sahar_abdelnabi , @nandofioretto, @saadu_ai, Shlomo Zilberstein, @ebagdasa 📄 Paper: arxiv.org/abs/2602.15198
🚨 Moltbook has shown significant vulnerabilities and safety risks when deploying multi-agent systems at scale, where AI agents can freely interact and coordinate with each other. 🚨 One potentially catastrophic risk is collusion where agents may undesirably coordinate to achieve a secondary objective. A large group of colluding agents can have devastating effects on the multi-agent system by influencing other agents' beliefs, actions, and propagating that influence through the network. But we don't have a sufficient way to audit these systems, specifically identifying collusive behavior of LLMs. 📄 We present our new arXiv paper: Colosseum: Auditing Collusion in Cooperative Multi-Agent Systems (arxiv.org/abs/2602.15198) What’s Colosseum? 🔍⚔️ A framework to audit collusive behavior in cooperative agentic multi-agent systems by grounding coordination in a DCOP and measuring collusion as regret vs. the cooperative optimum. Our framework can identify three collusion categories: 🤝Direct collusion — explicit coordination with realized collusive actions 🕵️‍♂️Attempted collusion — agents try/plan to collude in text but don’t successfully change actions/outcomes 🎭Hidden collusion — collusive outcomes without obvious/explicit signals (covert coordination) We stress-test collusion across: 🎯 objective misalignment 🗣️ persuasion tactics 🕸️ network influence 💡Key findings: 🕵️‍♂️ Emergent collusion: Many out-of-the-box models show a propensity to collude, despite not being prompted, when a secret side channel is added. 📝 We also find “collusion on paper”: agents plan to collude in text, but often take non-collusive actions. #tech #Agents #Moltbook #LLMs #AI #AiSafety
2
6
11
1,336
Saswat Das @ NeurIPS 2026 (ATL) retweeted
Hot take: the biggest risk in multi-agent systems isn’t one agent going rogue, it’s a whole swarm syncing up on the wrong goal. 🚨 In our latest work, we study how collusion can emerge once agents can freely interact and coordinate at scale, shaping other agents’ beliefs/actions and spreading influence through the network. The uncomfortable part: we still don’t have solid, standardized ways to audit collusive behavior in LLM-based multi-agent systems. 📄 Our new arXiv paper : Colosseum: Auditing Collusion in Cooperative Multi-Agent Systems arxiv.org/abs/2602.15198 What’s Colosseum? 🔍⚔️ A framework to audit collusion in cooperative agentic systems by grounding coordination in a DCOP and measuring collusion as regret vs. the cooperative optimum. We use Colosseum to detect 3 flavors of collusion: 🤝 Direct collusion — explicit coordination + realized collusive actions 🕵️‍♂️ Attempted collusion — agents plot in text but don’t shift actions/outcomes 🎭 Hidden collusion — collusive outcomes with no obvious signals (covert coordination) We stress-test across: 🎯 objective misalignment 🗣️ persuasion tactics 🕸️ network influence 💡 Two findings that stuck with us: 🕵️‍♂️ Emergent collusion: many out-of-the-box models start colluding without prompting when a secret side channel is introduced. 📝 “Collusion on paper”: lots of collusion talk… but the actions don’t always follow. If you’re deploying agent teams in production, coordination risk needs to be a first-class safety concern—right alongside single-agent robustness. Happy to chat / answer questions.
🚨 Moltbook has shown significant vulnerabilities and safety risks when deploying multi-agent systems at scale, where AI agents can freely interact and coordinate with each other. 🚨 One potentially catastrophic risk is collusion where agents may undesirably coordinate to achieve a secondary objective. A large group of colluding agents can have devastating effects on the multi-agent system by influencing other agents' beliefs, actions, and propagating that influence through the network. But we don't have a sufficient way to audit these systems, specifically identifying collusive behavior of LLMs. 📄 We present our new arXiv paper: Colosseum: Auditing Collusion in Cooperative Multi-Agent Systems (arxiv.org/abs/2602.15198) What’s Colosseum? 🔍⚔️ A framework to audit collusive behavior in cooperative agentic multi-agent systems by grounding coordination in a DCOP and measuring collusion as regret vs. the cooperative optimum. Our framework can identify three collusion categories: 🤝Direct collusion — explicit coordination with realized collusive actions 🕵️‍♂️Attempted collusion — agents try/plan to collude in text but don’t successfully change actions/outcomes 🎭Hidden collusion — collusive outcomes without obvious/explicit signals (covert coordination) We stress-test collusion across: 🎯 objective misalignment 🗣️ persuasion tactics 🕸️ network influence 💡Key findings: 🕵️‍♂️ Emergent collusion: Many out-of-the-box models show a propensity to collude, despite not being prompted, when a secret side channel is added. 📝 We also find “collusion on paper”: agents plan to collude in text, but often take non-collusive actions. #tech #Agents #Moltbook #LLMs #AI #AiSafety
1
3
261
Saswat Das @ NeurIPS 2026 (ATL) retweeted
What can we learn about LLMs' collusive behavior? We propose Colosseum to evaluate LLMs in new environments grounded in DCOPs and measure both conversations and actions and whether agents "walk the talk" on colluding. See the thread by @MasonNaka :
🚨 Moltbook has shown significant vulnerabilities and safety risks when deploying multi-agent systems at scale, where AI agents can freely interact and coordinate with each other. 🚨 One potentially catastrophic risk is collusion where agents may undesirably coordinate to achieve a secondary objective. A large group of colluding agents can have devastating effects on the multi-agent system by influencing other agents' beliefs, actions, and propagating that influence through the network. But we don't have a sufficient way to audit these systems, specifically identifying collusive behavior of LLMs. 📄 We present our new arXiv paper: Colosseum: Auditing Collusion in Cooperative Multi-Agent Systems (arxiv.org/abs/2602.15198) What’s Colosseum? 🔍⚔️ A framework to audit collusive behavior in cooperative agentic multi-agent systems by grounding coordination in a DCOP and measuring collusion as regret vs. the cooperative optimum. Our framework can identify three collusion categories: 🤝Direct collusion — explicit coordination with realized collusive actions 🕵️‍♂️Attempted collusion — agents try/plan to collude in text but don’t successfully change actions/outcomes 🎭Hidden collusion — collusive outcomes without obvious/explicit signals (covert coordination) We stress-test collusion across: 🎯 objective misalignment 🗣️ persuasion tactics 🕸️ network influence 💡Key findings: 🕵️‍♂️ Emergent collusion: Many out-of-the-box models show a propensity to collude, despite not being prompted, when a secret side channel is added. 📝 We also find “collusion on paper”: agents plan to collude in text, but often take non-collusive actions. #tech #Agents #Moltbook #LLMs #AI #AiSafety
1
3
11
824
Saswat Das @ NeurIPS 2026 (ATL) retweeted
The last few weeks, more than ever, tells us that the future is multi-agent 🚀 Collusion 🥷is a significant challenge in these systems, but we don't have frameworks and environments to audit and study it. Introducing Colosseum!! ⚔️
🚨 Moltbook has shown significant vulnerabilities and safety risks when deploying multi-agent systems at scale, where AI agents can freely interact and coordinate with each other. 🚨 One potentially catastrophic risk is collusion where agents may undesirably coordinate to achieve a secondary objective. A large group of colluding agents can have devastating effects on the multi-agent system by influencing other agents' beliefs, actions, and propagating that influence through the network. But we don't have a sufficient way to audit these systems, specifically identifying collusive behavior of LLMs. 📄 We present our new arXiv paper: Colosseum: Auditing Collusion in Cooperative Multi-Agent Systems (arxiv.org/abs/2602.15198) What’s Colosseum? 🔍⚔️ A framework to audit collusive behavior in cooperative agentic multi-agent systems by grounding coordination in a DCOP and measuring collusion as regret vs. the cooperative optimum. Our framework can identify three collusion categories: 🤝Direct collusion — explicit coordination with realized collusive actions 🕵️‍♂️Attempted collusion — agents try/plan to collude in text but don’t successfully change actions/outcomes 🎭Hidden collusion — collusive outcomes without obvious/explicit signals (covert coordination) We stress-test collusion across: 🎯 objective misalignment 🗣️ persuasion tactics 🕸️ network influence 💡Key findings: 🕵️‍♂️ Emergent collusion: Many out-of-the-box models show a propensity to collude, despite not being prompted, when a secret side channel is added. 📝 We also find “collusion on paper”: agents plan to collude in text, but often take non-collusive actions. #tech #Agents #Moltbook #LLMs #AI #AiSafety
1
4
23
2,503
Saswat Das @ NeurIPS 2026 (ATL) retweeted
Colosseum: Auditing Collusion in Cooperative Multi-Agent Systems Mason Nakamura, Abhinav Kumar, Saswat Das, Sahar Abdelnabi, Saaduddin Mahmud, Ferdinando Fioretto, Shlomo Zilberstein, Eugene Bagdasarian arxiv.org/abs/2602.15198 [𝚌𝚜.𝙼𝙰 𝚌𝚜.𝙰𝙸 𝚌𝚜.𝙲𝙻]
1
3
133
Saswat Das @ NeurIPS 2026 (ATL) retweeted
🚨One (more!) fully-funded PhD position in our group at Imperial College London – Privacy & Machine Learning 🔐🤖 starting Oct 2025 Plz RT 🔄
1
13
19
2,731
Saswat Das @ NeurIPS 2026 (ATL) retweeted
Our next talk is on Wednesday March 5th, jointly by @Joonhyuk_ko_03 and @WatIsDas, both students of @nandofioretto at UVA! More information can be found here: toc4fairness.org/toc4fairnes… See you next week!
1
1
579
Saswat Das @ NeurIPS 2026 (ATL) retweeted
I am hiring for a Ph.D. student to work in the areas of social network analysis, algorithms and fair machine learning. Please apply and join our highly motivated team. For more information please see the link to call in the tweet below.
3
53
1
138
16,709
Saswat Das @ NeurIPS 2026 (ATL) retweeted
ELLIS PhD position open in my group @FCAI_fi and upcoming ELLIS Institute Finland in differentially private machine learning in collaboration with Nokia Bell Labs. Apply here by 15 Nov: ellis.eu/news/ellis-phd-prog…
8
10
1,334
Saswat Das @ NeurIPS 2026 (ATL) retweeted
We’re excited to announce that the AAAI Workshop on Privacy-Preserving Artificial Intelligence (PPAI-25) is back! 🔗 ppai-workshop.github.io/ Please share widely! 🙏 Hosted at the 39th AAAI Conference on Artificial Intelligence (AAAI-25), this workshop will bring together AI researchers to explore and discuss current challenges and innovative solutions in privacy-preserving machine learning and related areas. Mark your calendars for **March 3, 2025**, in Philadelphia, USA! This year, emphasis will be placed on: - policy considerations and legal frameworks for privacy; - broader implications of privacy in AI agents and large language models (LLMs); - the societal impact of privacy within AI. For example, we’ll explore pressing questions like: “As embodied robotics agents become more integrated into our lives, how should we regulate them? If these agents have the capacity to see and hear us, could they ‘spy’ on us? What regulatory measures should be in place?” We invite submissions of technical papers and extended abstracts. The main deadline for submissions is *November 14 2024*. We also have a NeurIPS/AAAI Fast Track for rejected AAAI papers with average scores of at least 4.5; this fast track will have a lighter review process and has a deadline of *December 11th 2024*. Further information about the workshop, submission guidelines, and links for registration can be found here: ppai-workshop.github.io/ We look forward to your excellent submissions and engaging discussions at PPAI-25! Organizing committee: @JubaZiani @sistinerr, Jeremy Seeman, and Nando Fioretto.
2
28
1
69
10,351
Saswat Das @ NeurIPS 2026 (ATL) retweeted
Interested to work in a fast-paced research environment on the intersection of privacy and ML/AI (LLMs, diffusion models, and beyond)? Come join us 👩‍🎓👨‍🎓! Our research group at @imperialcollege has openings for fully-funded PhD students (Nov 15 deadline). Plz RT 🔁
1
4
2
10
1,659
Saswat Das @ NeurIPS 2026 (ATL) retweeted
Jailbreaks have become a new sort of ImageNet competition instead of helping us better understand LLM security. I wrote a blogpost about what I think valuable research could look like 🧵
7
28
12
174
52,168
Saswat Das @ NeurIPS 2026 (ATL) retweeted
The #AAAI Privacy Preserving AI workshop (#PPAI) is happening again!✨ In Philly on 3/3/25 ppai-workshop.github.io Stay tuned for updates, but for the moment, I hope you submit your work on technical/practical Privacy, privacy & LLMs Agents, and AI/privacy governance!
1
16
34
4,119
Saswat Das @ NeurIPS 2026 (ATL) retweeted
I’m recruiting PhD students interested in trustworthy ML and ML for security this cycle! Please reach out via email anytime or find me at @acm_ccs next week. Retweets are greatly appreciated!
4
63
1
141
21,882
Saswat Das @ NeurIPS 2026 (ATL) retweeted
Excited to share our #NeurIPS2024 work on combining diffusion models with constrained optimization to generate data adhering to constraints and physical principles (with guarantees!). Led by the amazing Jacob Christopher and with Stephen Baek. > arxiv.org/abs/2402.03559
4
53
2
308
31,265