1/29 New papers

[AmberLJC]

https://arxiv.org/abs/2601.17654
https://arxiv.org/abs/2601.20309 MLSys 26
https://arxiv.org/pdf/2510.22101 MLSys 26
https://arxiv.org/html/2506.05871v1
https://arxiv.org/pdf/2507.05411 MLSys 26
https://arxiv.org/pdf/2601.20408 MLSys 26

@claude

[claude]

Claude finished @AmberLJC's task in 2m 22s —— View job

---

Adding New Papers

I'm unable to fetch paper details from arXiv because I don't have web access permissions.

To complete this task, I have two options:

Option 1 (Recommended): Grant me web access by allowing WebFetch or curl commands, then I can:

• Automatically fetch all 6 paper titles and abstracts
• Properly categorize them into the correct sections (Training/Serving/Multi-Modal/etc.)
• Add properly formatted entries following the repository's style guide

Option 2: You can provide the paper titles and categories here in a comment, and I'll immediately add them to README.md in the correct format.

Papers waiting to be added:

1. 2601.17654
2. 2601.20309 (MLSys' 26)
3. 2510.22101 (MLSys' 26)
4. 2506.05871
5. 2507.05411 (MLSys' 26)
6. 2601.20408 (MLSys' 26)

Let me know how you'd like to proceed!

[AmberLJC]

@claude

1. Kareus: Joint Reduction of Dynamic and Static Energy in Large Model Training — https://arxiv.org/abs/2601.17654 ([arXiv][1])

2. SuperInfer: SLO-Aware Rotary Scheduling and Memory Management for LLM Inference on Superchips — https://arxiv.org/abs/2601.20309 ([arXiv][2])

3. Scaling Up Efficient Small Language Models Serving and Deployment for Semantic Job Search — https://arxiv.org/pdf/2510.22101 ([arXiv][3])

4. BestServe: Serving Strategies with Optimal Goodput in Collocation and Disaggregation Architectures — https://arxiv.org/html/2506.05871v1 ([arXiv][4])

5. AXLearn: Modular Large Model Training on Heterogeneous Infrastructure — https://arxiv.org/pdf/2507.05411 ([arXiv][5])

6. Meeting SLOs, Slashing Hours: Automated Enterprise LLM Optimization with OptiKIT — https://arxiv.org/pdf/2601.20408 ([arXiv][6])

[AmberLJC]

@claude

[claude]

Claude finished @AmberLJC's task in 1m 43s —— View job • claude/issue-51-20260130-0722 • Create PR ➔

---

Adding New Papers

• Read README.md to understand current structure
• Categorize papers into appropriate sections
• Add Kareus paper (training energy optimization)
• Add SuperInfer paper (LLM inference scheduling)
• Add Scaling Up Small LMs paper (serving/deployment)
• Add BestServe paper (serving strategies)
• Add AXLearn paper (model training)
• Add OptiKIT paper (enterprise LLM optimization)
• Commit and push changes
• Provide PR creation link

Summary

Successfully added all 6 papers to the README.md file:

Training (Pre-training section):

• Kareus: Joint Reduction of Dynamic and Static Energy in Large Model Training
• AXLearn: Modular Large Model Training on Heterogeneous Infrastructure | MLSys' 26

Serving (LLM serving section):

• SuperInfer: SLO-Aware Rotary Scheduling and Memory Management for LLM Inference on Superchips | MLSys' 26
• Scaling Up Efficient Small Language Models Serving: Serving and Deployment for Semantic Job Search | MLSys' 26
• BestServe: Serving Strategies with Optimal Goodput in Collocation and Disaggregation Architectures
• OptiKIT: Meeting SLOs, Slashing Hours - Automated Enterprise LLM Optimization | MLSys' 26

All entries follow the repository's formatting guidelines with proper arXiv URLs and venue information.

--- | Branch