Ehsan Kamalloo (@ehsk0) profile picture

Ehsan Kamalloo

@ehsk0

Research Scientist @ServiceNowRSRCH

ehsk.github.io
Joined August 2013
  • Tweets253
  • Following620
  • Followers354
9Media
Photo by Ehsan Kamalloo (@ehsk0)Photo by Ehsan Kamalloo (@ehsk0)Photo by Ehsan Kamalloo (@ehsk0)Photo by Ehsan Kamalloo (@ehsk0)Photo by Ehsan Kamalloo (@ehsk0)Photo by Ehsan Kamalloo (@ehsk0)Photo by Ehsan Kamalloo (@ehsk0)Photo by Ehsan Kamalloo (@ehsk0)Photo by Ehsan Kamalloo (@ehsk0)
Still in 🇭🇺 for the final day of @emnlpmeeting? Join us in Room P1 for the 🤖 REALM 🤖 workshop!! We kicked off the session with a fascinating keynote on "Building Code Models" by @b_roziere from @MistralAI #EMNLP2026 #LLMAgents #NLProc
Image from Akhil Arora @EMNLP26 🇭🇺 (@aroraakhilcs)
Image from Akhil Arora @EMNLP26 🇭🇺 (@aroraakhilcs)
Image from Akhil Arora @EMNLP26 🇭🇺 (@aroraakhilcs)
Akhil Arora @EMNLP26 🇭🇺 (@aroraakhilcs) profile pictureAkhil Arora @EMNLP26 🇭🇺@aroraakhilcsOct 22
📣 🤖 REALM 🤖 is happening next Thursday @emnlpmeeting in Budapest! 🇭🇺 📅 Oct 29 · Room P1 · from 09:15 (hybrid) 📊 264 submissions, 163 accepted: 6 orals 🎤, 9 spotlights ✨, 148 posters🪧 🎤 6 keynotes, schedule ⬇️ realm-workshop.github.io/sch… #EMNLP2026 #LLMAgents #NLProc
1
2
9
513
An exciting lineup of speakers, papers, and discussions on the latest advances and challenges in AI agents 🤖 🚀 Join us next Thursday at #REALM2026 in Budapest! #EMNLP2026
Akhil Arora @EMNLP26 🇭🇺 (@aroraakhilcs) profile pictureAkhil Arora @EMNLP26 🇭🇺@aroraakhilcsOct 22
📣 🤖 REALM 🤖 is happening next Thursday @emnlpmeeting in Budapest! 🇭🇺 📅 Oct 29 · Room P1 · from 09:15 (hybrid) 📊 264 submissions, 163 accepted: 6 orals 🎤, 9 spotlights ✨, 148 posters🪧 🎤 6 keynotes, schedule ⬇️ realm-workshop.github.io/sch… #EMNLP2026 #LLMAgents #NLProc
1
7
132
Ehsan Kamalloo retweeted
Nicolas Gontier (@nicogontier) profile picture
Nicolas Gontier@nicogontierOct 22
If you are at EMNLP next week, be sure to checkout our REALM workshop! Some incredible agentic work will be presented 🤖
Akhil Arora @EMNLP26 🇭🇺 (@aroraakhilcs) profile pictureAkhil Arora @EMNLP26 🇭🇺@aroraakhilcsOct 22
📣 🤖 REALM 🤖 is happening next Thursday @emnlpmeeting in Budapest! 🇭🇺 📅 Oct 29 · Room P1 · from 09:15 (hybrid) 📊 264 submissions, 163 accepted: 6 orals 🎤, 9 spotlights ✨, 148 posters🪧 🎤 6 keynotes, schedule ⬇️ realm-workshop.github.io/sch… #EMNLP2026 #LLMAgents #NLProc
2
5
194
⏳ Final 24 hours to commit your ARR-reviewed papers to the @emnlpmeeting REALM Workshop! We welcome contributions on AI agents 🤖 — archival or non-archival. 📅 ARR commitment Deadline: Aug 31, 23:59 AoE 📰 CfP: lnkd.in/guj-Jr5y #EMNLP2026 #ARR #AIAgents #LLMAgents

LinkedIn

This link will take you to a page that’s not on LinkedIn

lnkd.in
Akhil Arora @EMNLP26 🇭🇺 (@aroraakhilcs) profile pictureAkhil Arora @EMNLP26 🇭🇺@aroraakhilcsJun 3
AI agents like @openclaw 🦞 are everywhere, answering emails, managing calendars, doing our chores for us 📣 REALM is back for year 2! Workshop for Research on Agent Language Models at #EMNLP2026, Budapest 🇭🇺 Stellar lineup ⬇️ 📅 Submit by July 17, 23:59 AoE #LLMAgents #NLProc
REALM 2026 workshop invited speakers and panelists: Maria Brbić (EPFL), Vishrav Chaudhary (Meta Superintelligence Labs), Marc-Alexandre Côté (Microsoft Research), Manling Li (Northwestern University), Baptiste Rozière (Mistral AI), Graham Neubig (Carnegie Mellon University), Perouz Taslakian (ServiceNow AI Research).

ALT REALM 2026 workshop invited speakers and panelists: Maria Brbić (EPFL), Vishrav Chaudhary (Meta Superintelligence Labs), Marc-Alexandre Côté (Microsoft Research), Manling Li (Northwestern University), Baptiste Rozière (Mistral AI), Graham Neubig (Carnegie Mellon University), Perouz Taslakian (ServiceNow AI Research).

1
2
358
With EMNLP results behind us, good news or bad, we at REALM welcome your contribution (archival or non-archival) on AI agents🤖. 📅 ARR commitment: August 31, 2026 📰 CfP: realm-workshop.github.io/cal… @aclmeeting #EMNLP2026 #ARR #NLProc #AIAgents #LLMAgents
Akhil Arora @EMNLP26 🇭🇺 (@aroraakhilcs) profile pictureAkhil Arora @EMNLP26 🇭🇺@aroraakhilcsJun 3
AI agents like @openclaw 🦞 are everywhere, answering emails, managing calendars, doing our chores for us 📣 REALM is back for year 2! Workshop for Research on Agent Language Models at #EMNLP2026, Budapest 🇭🇺 Stellar lineup ⬇️ 📅 Submit by July 17, 23:59 AoE #LLMAgents #NLProc
REALM 2026 workshop invited speakers and panelists: Maria Brbić (EPFL), Vishrav Chaudhary (Meta Superintelligence Labs), Marc-Alexandre Côté (Microsoft Research), Manling Li (Northwestern University), Baptiste Rozière (Mistral AI), Graham Neubig (Carnegie Mellon University), Perouz Taslakian (ServiceNow AI Research).

ALT REALM 2026 workshop invited speakers and panelists: Maria Brbić (EPFL), Vishrav Chaudhary (Meta Superintelligence Labs), Marc-Alexandre Côté (Microsoft Research), Manling Li (Northwestern University), Baptiste Rozière (Mistral AI), Graham Neubig (Carnegie Mellon University), Perouz Taslakian (ServiceNow AI Research).

3
6
380
Ehsan Kamalloo retweeted
Victor Zhong (@hllo_wrld) profile picture
Victor Zhong@hllo_wrldAug 21
Just when I thought things couldn’t get worse.. student fought hard during rebuttal and managed to improve score, with a AC recommendation for EMNLP - only to get rejected by EMNLP itself.
Victor Zhong (@hllo_wrld) profile pictureVictor Zhong@hllo_wrldJul 15
This ARR is the 1st time I'm having serious doubt about the state of scientific review. I am now increasingly negative about an "open internet" and "open research" in the sense that anybody can put out anything out and the system works because "most" things are well-intentioned.
7
1
40
7,223
Ehsan Kamalloo retweeted
Rafa Pardinas (@muchomuchacho) profile picture
Rafa Pardinas
@muchomuchachoJul 9
Our paper Apriel-OpenReasoner: RL Post-Training for General-Purpose and Efficient Reasoning has been accepted to COLM 2026! Apriel-OpenReasoner matches or beats comparable open models on AIME 2025, GPQA, MMLU-Pro & LiveCodeBench with 30-50% shorter traces.
Image from Rafa Pardinas (@muchomuchacho)
1
3
1
17
2,303
Ehsan Kamalloo retweeted
Pouya Sadeghi (@Ipouyall) profile picture
Pouya Sadeghi@IpouyallJul 8
🏆 Thrilled that our ACL 2026 paper “To Lie or Not to Lie?” was selected as an ACL 2026 SAC Highlight! 🎉 The work asks a simple but urgent question: are LLMs equally safe around the world? 🧵👇 @aclmeeting #ACL2026NLP #NLProc
Image from Pouya Sadeghi (@Ipouyall)
6
17
1
67
4,257
Hey :) I'm looking for 5 emergency reviewers for ARR submissions🚨📷Broad Topics: Reasoning, Culture, Evals, MAS, Alignment, Reliability Reviews due in 3 days: Tue 7 Jul 23:59 AoE 🙃. If you are interested, please DM or email me! #NLProc #NLP #LLMs #ACL #ARR
1
3
8
1,411
Ehsan Kamalloo retweeted
Maryam Hashemzadeh (@MaryamHashemz) profile picture
Maryam Hashemzadeh
@MaryamHashemzJun 4
So excited to co-organize the Research on Agent Language Models (REALM) workshop at #EMNLP2026! 🚀 We are looking forward to some fantastic discussions and an amazing program. 📅 Deadline: July 17 🌐 Details: realm-workshop.github.io
Akhil Arora @EMNLP26 🇭🇺 (@aroraakhilcs) profile pictureAkhil Arora @EMNLP26 🇭🇺@aroraakhilcsJun 3
AI agents like @openclaw 🦞 are everywhere, answering emails, managing calendars, doing our chores for us 📣 REALM is back for year 2! Workshop for Research on Agent Language Models at #EMNLP2026, Budapest 🇭🇺 Stellar lineup ⬇️ 📅 Submit by July 17, 23:59 AoE #LLMAgents #NLProc
REALM 2026 workshop invited speakers and panelists: Maria Brbić (EPFL), Vishrav Chaudhary (Meta Superintelligence Labs), Marc-Alexandre Côté (Microsoft Research), Manling Li (Northwestern University), Baptiste Rozière (Mistral AI), Graham Neubig (Carnegie Mellon University), Perouz Taslakian (ServiceNow AI Research).

ALT REALM 2026 workshop invited speakers and panelists: Maria Brbić (EPFL), Vishrav Chaudhary (Meta Superintelligence Labs), Marc-Alexandre Côté (Microsoft Research), Manling Li (Northwestern University), Baptiste Rozière (Mistral AI), Graham Neubig (Carnegie Mellon University), Perouz Taslakian (ServiceNow AI Research).

1
4
323
Ehsan Kamalloo retweeted
Nikolai Rozanov (@ai_nikolai) profile picture
Nikolai Rozanov@ai_nikolaiJun 3
We have an amazing line-up of speakers for the 2nd Edition of the Agent workshop (and amazing organisers too). Submission is now open. See the post below. @Cote_Marc @ManlingLi_ @gneubig @vishrav @mariabrbic @b_roziere @PerouzT
Akhil Arora @EMNLP26 🇭🇺 (@aroraakhilcs) profile pictureAkhil Arora @EMNLP26 🇭🇺@aroraakhilcsJun 3
AI agents like @openclaw 🦞 are everywhere, answering emails, managing calendars, doing our chores for us 📣 REALM is back for year 2! Workshop for Research on Agent Language Models at #EMNLP2026, Budapest 🇭🇺 Stellar lineup ⬇️ 📅 Submit by July 17, 23:59 AoE #LLMAgents #NLProc
REALM 2026 workshop invited speakers and panelists: Maria Brbić (EPFL), Vishrav Chaudhary (Meta Superintelligence Labs), Marc-Alexandre Côté (Microsoft Research), Manling Li (Northwestern University), Baptiste Rozière (Mistral AI), Graham Neubig (Carnegie Mellon University), Perouz Taslakian (ServiceNow AI Research).

ALT REALM 2026 workshop invited speakers and panelists: Maria Brbić (EPFL), Vishrav Chaudhary (Meta Superintelligence Labs), Marc-Alexandre Côté (Microsoft Research), Manling Li (Northwestern University), Baptiste Rozière (Mistral AI), Graham Neubig (Carnegie Mellon University), Perouz Taslakian (ServiceNow AI Research).

2
1
6
6,720
Ehsan Kamalloo retweeted
Nicolas Gontier (@nicogontier) profile picture
Nicolas Gontier@nicogontierJun 3
The second edition of our agent workshop is now accepting submissions! Check out our website for the full details! realm-workshop.github.io/cal…
Akhil Arora @EMNLP26 🇭🇺 (@aroraakhilcs) profile pictureAkhil Arora @EMNLP26 🇭🇺@aroraakhilcsJun 3
AI agents like @openclaw 🦞 are everywhere, answering emails, managing calendars, doing our chores for us 📣 REALM is back for year 2! Workshop for Research on Agent Language Models at #EMNLP2026, Budapest 🇭🇺 Stellar lineup ⬇️ 📅 Submit by July 17, 23:59 AoE #LLMAgents #NLProc
REALM 2026 workshop invited speakers and panelists: Maria Brbić (EPFL), Vishrav Chaudhary (Meta Superintelligence Labs), Marc-Alexandre Côté (Microsoft Research), Manling Li (Northwestern University), Baptiste Rozière (Mistral AI), Graham Neubig (Carnegie Mellon University), Perouz Taslakian (ServiceNow AI Research).

ALT REALM 2026 workshop invited speakers and panelists: Maria Brbić (EPFL), Vishrav Chaudhary (Meta Superintelligence Labs), Marc-Alexandre Côté (Microsoft Research), Manling Li (Northwestern University), Baptiste Rozière (Mistral AI), Graham Neubig (Carnegie Mellon University), Perouz Taslakian (ServiceNow AI Research).

1
5
324
📢REALM returns for year 2 to dig into the questions that actually matter now that agents are everywhere: reliability, safety, long-horizon planning, multi-agent systems & more. We'd love to see your work! Submit by 📅 Jul 17. Join us Oct 29 at #EMNLP2026 in Budapest! #AIagents
Akhil Arora @EMNLP26 🇭🇺 (@aroraakhilcs) profile pictureAkhil Arora @EMNLP26 🇭🇺@aroraakhilcsJun 3
AI agents like @openclaw 🦞 are everywhere, answering emails, managing calendars, doing our chores for us 📣 REALM is back for year 2! Workshop for Research on Agent Language Models at #EMNLP2026, Budapest 🇭🇺 Stellar lineup ⬇️ 📅 Submit by July 17, 23:59 AoE #LLMAgents #NLProc
REALM 2026 workshop invited speakers and panelists: Maria Brbić (EPFL), Vishrav Chaudhary (Meta Superintelligence Labs), Marc-Alexandre Côté (Microsoft Research), Manling Li (Northwestern University), Baptiste Rozière (Mistral AI), Graham Neubig (Carnegie Mellon University), Perouz Taslakian (ServiceNow AI Research).

ALT REALM 2026 workshop invited speakers and panelists: Maria Brbić (EPFL), Vishrav Chaudhary (Meta Superintelligence Labs), Marc-Alexandre Côté (Microsoft Research), Manling Li (Northwestern University), Baptiste Rozière (Mistral AI), Graham Neubig (Carnegie Mellon University), Perouz Taslakian (ServiceNow AI Research).

1
211
Ehsan Kamalloo retweeted
Kusha Sareen (@KushaSareen) profile picture
Kusha Sareen@KushaSareenMay 13
Can LLMs adapt continually without losing base skills? Fast-Slow Training (FST) pairs "slow" weights with "fast" context. FST vs. RL: • 3x more sample-efficient • Higher performance ceiling • Less KL drift (better plasticity) • Continual learning: succeeds where RL stalls
Image from Kusha Sareen (@KushaSareen)
20
92
12
548
139,171
Ehsan Kamalloo retweeted
Rafa Pardinas (@muchomuchacho) profile picture
Rafa Pardinas
@muchomuchachoMay 8
Our first vLLM V0→V1 run on PipelineRL looked broken. @ehsk0 and I almost reached for an objective-side correction. That would have been the wrong fix. The real problem: four mismatches in the rollout backend. 🧵
Image from Rafa Pardinas (@muchomuchacho)
1
4
1
7
2,634
Ehsan Kamalloo retweeted
Alexandre Lacoste (@alex_lacoste_) profile picture
Alexandre Lacoste@alex_lacoste_Apr 27
Research institutions weren't built for a world where AI can generate papers for $15, conferences get 20k+ submissions, and a handful of closed labs are pulling ahead fast. I wrote about what the research community could do about it, and proposed a concrete system to help. The Research Commons: multi-blind coordination, anonymous channels, roles for requesters/executors/reproducers/reviewers, and attribution that actually reflects who did what. alexlacoste.substack.com/p/t…
Image from Alexandre Lacoste (@alex_lacoste_)
5
8
2
26
2,402
(1/8) 🚀 Introducing Super Apriel: One Checkpoint, Many Speeds Train once → serve at any speed-quality tradeoff We release: ✓ 15B supernets with 4 mixers/layer ✓ Training code (Fast-LLM) ✓ vLLM serving extension 🧵 How it works ↓
Image from ServiceNow AI Research (@ServiceNowRSRCH)
1
10
6
32
10,941
Ehsan Kamalloo retweeted
Rafa Pardinas (@muchomuchacho) profile picture
Rafa Pardinas
@muchomuchachoApr 19
Cross-domain RL training works and PipelineRL now supports it natively. We also incorporate adaptive domain sampling to keep the sampling proportions on target throughout training. Evidence from our recent paper: arxiv.org/abs/2604.02007 code: github.com/ServiceNow/Pipeli…
Apriel-Reasoner: RL Post-Training for General-Purpose and...

Apriel-Reasoner: RL Post-Training for General-Purpose and...

Building general-purpose reasoning models using reinforcement learning with verifiable rewards (RLVR) across diverse domains has been widely adopted by frontier open-weight models. However, their...

arxiv.org
1
6
153
Ehsan Kamalloo retweeted
Rafa Pardinas (@muchomuchacho) profile picture
Rafa Pardinas
@muchomuchachoApr 9
Better reasoning does not have to mean longer reasoning. Apriel OpenReasoner: fully reproducible multi-domain RL post-training using public datasets. 30-50% shorter traces, no quality trade-off. @ServiceNowRSRCH @ehsk0 @dvazquezcv @alexandredrouin
4
5
2
12
3,490
Ehsan Kamalloo retweeted
Patrice Bechard (@patricebechard) profile picture
Patrice Bechard
@patricebechardApr 2
What if we didn't need MCP servers after all? What if we didn't need browser-use agents either? What if... Claude Code was enough? In our latest paper, we test exactly this: Can simple terminal agents outperform web agents and tool-based agents on real enterprise tasks?
2
17
3
37
7,602
Load more