New AI Models (Last 6 Months) — Page 427

Qwen2.5-Coder-CWS-MCEVALHARD-7B-Base

qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step450

qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step200

qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step250

ee_gol_grp_f1_form_spanOver

qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step550

ee_gol_grp_f1_form_wo_ns

PureRL-7B-v5-13-fmt025-accW15

qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step250

qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step200

qwen2.5-0.5b-grpo-arithmetic

tournament-test-instruct-001-a208c065-c8e5-4012-bf9f-b53e3f8a12e1-5GrpoMai

Llama-3.1-8B-Instruct_grpo_ppl_adv_rollout_8_kl_0.001_20260516_140637_step290

Llama_3.1_8B_Instruct_grpo_ppl_adv_step580

rudolph-v1-merged

Qwen3-1.7B-icl-100shot-id_only

safety_model

PureRL-1.5B-v5-06-mc

Affine-top15-5ELt9A1qzud3e8hKJDEXun9nFjydoYy4hagq52xcjNGcKrEm

PureRL-1.5B-v9F-digit-w100

affine-champ-clone-5Ct6ocEEjf59tak3RyhsetcfAtAyFL5e6SEXSvzxMryrgMK3

qwen3-1.7B-sft-instruct-ckpt350

volta-energy-parser

Qwen2.5-Coder-CONTROL-MCEVALHARD-1.5B-Base-1

affine-pathc-v7-champ-5Hg5ggzJCDeQgqs2h7fQfCGkfDUEbcHo5rEbmSEwBmnQXG8X

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

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

PureRL-1.5B-v7-stage1-qa-instruct

tofu_Llama-3.2-1B-Instruct_forget10_NPO_qat-int4

qwen2.5-3b-medpt-lora

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

math_model

GRPO-Model

general_knowledge_model

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

Gemma-3-4B-IT-HI-SynthDolly-r16alpha128-E5-S73

Perovskite-RL

Qwen3-8B-v1-test

math_no_think_8_qwen3_4b_instruct_sft

sac-gspo-cl3e3-drgrpo-llama32-3b-deepscaler-step841-best-pass1-15.21-8xH200

opd_medical_qwen3-0.6b_frozen_teacher_forward_kl

qwen2.5-0.5b-instruct-openai-gsm8k-dppo-topk