New AI Models (Last Year) — Page 611

gemma-2-9b-it-lr3e-5-safeinstr-lr1e-5-0.05

safety_model

llama-3.1-tulu-8b-dpo-abstention

teacher_grpo_forward_kl-1_16k

cb-wmdp-Llama-3.1-8B-Instruct-bfbf3e38793c

Qwen3-14B-pragrest-outcome-0.8-qa-only-kl-0.02-lr-4e-6-2-no-easy-no-hard-vanilla-sft_step_16

qwen3-4b-thinking-grpo-strict70

Qwen2.5-Math-1.5B_grpo_entropy_rollout_8_ent_0.001_USE_KL_0.001_resume_20260512_222805_step580

RLCR-1.5B-hotpot-rac-lr5e6

Qwen3-0.6B-Reverse-Text-SFT

Qwen2.5-Math-7B_grpo_entropy_rollout_8_ent_0.001_USE_KL_0.001_20260513_122028_step580

Affine-swe3-5Fn18zy4SEBEKjYeWVB92hR8ZCxxK1c4p2jPvbRH2bfpQTXT

qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step150

qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step300

qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_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

gORM-14B-2-merged

qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step50

ee_gol_grp_f1_form_wo_ns

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

qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step150

gORM-14B-3-merged

Qwen_base_asap_shot7_sft_fold0

qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step250

qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step200

multilingual_model

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

Qwen2.5-Math-1.5B_grpo_ppl_only_rollout_8_ent_0.0_kl_True_0.001_20260515_154012_step580

Llama-3.1-8B-Instruct_grpo_ppl_adv_rollout_8_kl_0.001_20260516_140637_step232

Qwen2.5-Math-7B_grpo_base_step580

gPRM-14B-3-merged

llama-8b-instruct-email-classify

HyperExtract-LLM

safety_model

RASA-all3-Phi-3.5-MoE-instruct

venue-model-merged

assn2-simpo-llama32-1b

safety_model

gPRM-14B-2-merged

Llama-3.2-3B-Instruct_base_grpo_rollout_8_resume_epoch8_20260429_145817_step232