SEA Reasoning Math Benchmarks These extended benchmarks were generated via ClimbSeq using aisingapore/Gemma-SEA-LION-v4-27B-IT. Towards equitable AI for all communities. sglim/aime25-SEA Viewer • Updated Aug 3 • 30 • 15 sglim/hmmt_nov_2025-SEA Viewer • Updated Aug 3 • 30 • 22
SEA Reasoning Language Models These models were post-trained from Qwen/Qwen3-VL-8B-Thinking using Onramp Sequence Cross-Distillation. Towards equitable AI for all communities. Native Multilingual Chain-of-Thought Reasoning in Low-Resource Southeast Asian Languages Paper • 2608.00533 • Published Aug 1 • 1 sglim/Qwen3-VL-8B-Thinking-SEA-Reasoning-Answer 9B • Updated Aug 5 • 12 • 1 sglim/Qwen3-VL-8B-Thinking-SEA-Reasoning 9B • Updated Aug 5 • 21 • 1
Native Multilingual Chain-of-Thought Reasoning in Low-Resource Southeast Asian Languages Paper • 2608.00533 • Published Aug 1 • 1
SEA Reasoning Math Benchmarks These extended benchmarks were generated via ClimbSeq using aisingapore/Gemma-SEA-LION-v4-27B-IT. Towards equitable AI for all communities. sglim/aime25-SEA Viewer • Updated Aug 3 • 30 • 15 sglim/hmmt_nov_2025-SEA Viewer • Updated Aug 3 • 30 • 22
SEA Reasoning Language Models These models were post-trained from Qwen/Qwen3-VL-8B-Thinking using Onramp Sequence Cross-Distillation. Towards equitable AI for all communities. Native Multilingual Chain-of-Thought Reasoning in Low-Resource Southeast Asian Languages Paper • 2608.00533 • Published Aug 1 • 1 sglim/Qwen3-VL-8B-Thinking-SEA-Reasoning-Answer 9B • Updated Aug 5 • 12 • 1 sglim/Qwen3-VL-8B-Thinking-SEA-Reasoning 9B • Updated Aug 5 • 21 • 1
Native Multilingual Chain-of-Thought Reasoning in Low-Resource Southeast Asian Languages Paper • 2608.00533 • Published Aug 1 • 1
