Models

qwen-hf-fewshot-iter-contam-np-iter1

SOR-ColdBrew-12B-Think-Base

goldengoose-gumbel_gmrel_tau1.00-25grp

brainalign-qwen2.5-1.5b-C

qwen-2.5-3b-roman-konkani-v3

qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step550

qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step500

qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step580

llama-3.1-8b-r512-gd-random

llama3-8b-legal-chatbot-grpo

llama-3.1-8b-r128-gd-random

augmented-0fc49138d5f71e66

PureRL-1.5B-v12A-lam002

PureRL-1.5B-v13C-lam010

Llama-3.1-8B-target-only-last-third

LlamaPlushie-3-8B-3

goldengoose-gumbel_gradsim_tau2.00-25grp

Zigroo-Mental_consultant2-merged

finch_8b_soft_without_held_out_expr_purpose_qwen_1.0e-5_1.0_train42_cosine

qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step450

qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step500

int_qwen3-4b_distill_teacher_reverse_kl_lr1e-7

FAME_PO_llama32-1b-10-instruct-qa

llama-3.1-8b-r128-gd-random-qres1

Qwen3-8B-EN-SynthDolly-r16alpha32-E3-S73

Qwen3-14B-EN-SynthDolly-r16alpha32-E8-S73

Llama-3.1-8B-Instruct-EN-SynthDolly-r16alpha32-E8-S73

CEEH_7B_ME

goldengoose-gumbel_gradsim_tau0.10-25grp

L3-CharThink-Base-Test

augmented-9628c62b4208063a

PrAg-PO-Qwen3-1.7b-step720

atlas-mini

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_step400

qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step200

Qwen2.5-7B

PureRL-1.5B-v12C-lam010

PureRL-1.5B-v12D-lam025

Llama-3.1-8B-bad-medical-top80

Llama-3.1-8B-good-vs-bad-last-third

Qwen3-8B-reward-hacks-top10