Long Papers
At most 8 pages long (excluding references and appendices), in LaTeX format using the ICLR 2026 conference style.
Machine Learning/CORE Unranked
SPOT 2026
The 1st Workshop on Scaling Post-training for LLMs
SPOT @ ICLR 2026 is a workshop focused on establishing a principled scientific framework for scaling post-training in large language models, encompassing supervised fine-tuning, reinforcement learning, and related methodologies. It brings together researchers to address underexplored challenges in efficiency, stability, and asymptotic performance across algorithms, systems, architectures, data, and safety.
Key deadlines
Verified September 9, 2026Full paper
February 7, 2026
AoE
Workshop timeline
Submission and decisions
Full paperKey deadline
February 7, 2026 · AoE
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Short/Tiny Papers
At most 4 pages long (excluding references and appendices), in LaTeX format using the ICLR 2026 conference style; authors may be earmarked for potential ICLR funding.
Research areas in scope
01
Topics of Interest
Algorithms: SFT and RL formulations, reward modeling, and objectives that enable stable and compute-efficient scalingSystems: Infrastructure challenges unique to post-training, including rollout latency, training–inference mismatches, distributed execution, and KV-cache managementArchitectures: The impact of design choices such as MoE, sparse attention, and model specialization on sample efficiency and generalizationData: The role of data curation and synthetic data in scaling post-training and shaping generalizationEnvironments & Safety: Post-training in multimodal, embodied, and real-world settings, with emphasis on evaluation fidelity and safetyFeedback & Verification: Efficient, dense feedback mechanisms as evaluation costs increasingly rival training costs
Policies worth checking twice
- Submissions must be previously unpublished.
- Concurrent submissions under review at other venues are welcome.
- Submissions must be anonymized for double-blind review.
- Accepted papers will appear on OpenReview but do not constitute formal proceedings.
- Authors remain free to submit and publish their work elsewhere.
- Authors must upload a camera-ready version by April 1, 2026, with a maximum of 1 additional page for revisions (7 pages for technical papers, 5 pages for position/vision papers).
- At least one author per submission is encouraged to serve as a reviewer.
- Posters must be in portrait format and not exceed 36in (H) x 24in (W) or 91cm (H) x 61cm (W).
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Last verified September 9, 2026