New AI Models (Last Year) — Page 583

Qwen3-1.7B-icl-3shot-dpo-replace_copy

math_model

grpo_adv_rollout_8_20260513_123609_USE_KL_True_step580

Qwen2.5-Math-1.5B_grpo_ppl_adv_rollout_8_ent_0.0_kl_True_0.001_20260515_153830_step580

Qwen3-1.7B-icl-100shot-id_only

grpo_rollout_8_step580

ee_gol_grpo_scratch_dapo_mcts

canoe-modified-2ep

canoe-1_1-270steps

Affine-ccc0-5EcVrCC1oFQPLeKoxTFpoPbBLQaNfooVRHSWZpPvrJBA6RxL

multilingual_model

safety_model

Qwen2.5-Coder-CONTROL-MCEVALHARD-1.5B-Base-6

multilingual_model

general_knowledge_model

affine-pathc-v7-champ-5Hg5ggzJCDeQgqs2h7fQfCGkfDUEbcHo5rEbmSEwBmnQXG8X

infinite-craft-model

gemma-3-4b-opt3-with-gt

multilingual_model

ipo-countdown-qwen2.5-0.5b

math_model

Llama-3.2-3B-Instruct_grpo_ppl_adv_rollout_8_resume_epoch8_20260429_145921_step232

PureRL-7B-v7-s2-async-l2-maskon

sac-gspo-cl3e3-drgrpo-r1distill-qwen1.5b-step420-aime24-34_3-temp1

Llama-3.2-3B-Instruct-DA-SynthDolly-r16alpha32-E8-S73

Qwen2.5-Coder-PROD-MCEVALHARD-1.5B-Base-5

group_model

GRPO-Model

Qwen2.5-Coder-OVERFIT-MCEVALHARD-1.5B-Base

audit-unlearn-npo-llama31-8b-dolly

math_no_think_8_qwen3_4b_instruct_sft

Graph_R1_RL

math_no_think_8_qwen3_4b_base_sft

ee_gol_grp_f1_form_over

sac-gspo-cl3e3-drgrpo-qwen25-math-1.5b-step1500

sac-gspo-cl3e3-drgrpo-r1distill-qwen1.5b-24k-temp1-step641

sllm-shady

sage-qwen3-4b-code-coevolve-gen-final

PureRL-1.5B-v7-s2-l2-kl-w1-b0

grpo_ppl_adv_rollout_8_step580

sac-gspo-cl3e3-drgrpo-r1distill-qwen1.5b-24k-temp1-step821-aime24-40pct

sac-gspo-cl3e3-drgrpo-r1distill-qwen1.5b-24k-temp1-step881-aime24-40pct