|
QLoRA on a code-repair dataset: eval loss rises after epoch 1 even with 2× more data. Early stopping, or mask the copied tokens?
|
|
1
|
14
|
November 2, 2026
|
|
Roadmap for AI learning preparation
|
|
2
|
110
|
November 2, 2026
|
|
An open research lab for measuring — not prompting — music production in a DAW
|
|
16
|
530
|
November 2, 2026
|
|
Freedream-1B-Instruct-V2-preview
|
|
1
|
52
|
November 2, 2026
|
|
Project Guardian: A Hybrid Safety System for AI Agents
|
|
6
|
90
|
November 1, 2026
|
|
Ifparse 2.0 Release
|
|
0
|
42
|
November 1, 2026
|
|
What are the biggest challenges you face when sourcing high-quality, domain-specific datasets for training and evaluating AI models?
|
|
3
|
116
|
November 2, 2026
|
|
The chat is not working
|
|
2
|
73
|
November 1, 2026
|
|
SCCN Unthink: semantic scene reconstruction across seven model/surface configurations
|
|
1
|
41
|
November 2, 2026
|
|
Can we prove this: the code submitted with this research paper actually produced the numbers reported?
|
|
14
|
264
|
October 28, 2026
|
|
When the fastest LLM reply skips the task: a proposed comparison protocol
|
|
6
|
78
|
November 1, 2026
|
|
Misalignment is not required for an agent to act without authorization
|
|
58
|
1209
|
November 2, 2026
|
|
How to evaluate deceptive dialogue in LLM-based NPCs?
|
|
11
|
291
|
November 2, 2026
|
|
We gave AI a trust we never gave people - The third axis, after capability and permission
|
|
21
|
751
|
November 2, 2026
|
|
ZeroGPU Space stuck in APP_STARTING after Git push — Factory reboot doesn't help
|
|
7
|
157
|
October 27, 2026
|
|
CAEL 0.3.7: non-neural assistant vs Qwen3-0.6B and Qwen3.5-0.8B on a frozen 80-item test
|
|
12
|
151
|
November 2, 2026
|
|
My First 1B transformer model trained from scratch has published
|
|
4
|
103
|
October 31, 2026
|
|
🚀 Introducing ZY-S2: A 1.5B Open-Source Coding Model by ZYR AI
|
|
1
|
61
|
November 2, 2026
|
|
The model sizes
|
|
1
|
42
|
November 1, 2026
|
|
A heads up on REDDIT for ordinary people
|
|
8
|
295
|
October 30, 2026
|
|
RAG eval that doesn’t lie to you: hit-rate first, then faithfulness
|
|
5
|
99
|
November 2, 2026
|
|
Random vs scaffold splits: single-seed baselines on 6 chemistry datasets
|
|
4
|
86
|
November 2, 2026
|
|
Best small open embedding model for semantic search over a Q&A forum archive (CPU-only)?
|
|
3
|
72
|
November 1, 2026
|
|
Memory is identity: it should select, forget, distort, and gradually build a self. Not just store and retrieve
|
|
22
|
288
|
October 30, 2026
|
|
Can you make an agent betray its owner's instructions by changing what it remembers?
|
|
6
|
82
|
November 1, 2026
|
|
OxCode: an AI software engineer for VS Code (82% SWE-bench Verified). Looking for feedback
|
|
0
|
35
|
October 31, 2026
|
|
Freedream-1B-Instruct V1.1 from scratch published!
|
|
0
|
43
|
October 31, 2026
|
|
Triton-hf-builder: Generate NVIDIA Triton repositories from Hugging Face checkpoints
|
|
2
|
50
|
October 31, 2026
|
|
What is the purpose of each Hugging Face tokenizer file, and what does tokenizer_config.json add beyond tokenizer.json?
|
|
2
|
50
|
October 30, 2026
|
|
Can ethical cooperation between humans and AI agents be tested?
|
|
22
|
333
|
November 2, 2026
|