Models

Qwen2.5-Math-1.5B_grpo_entropy_rollout_8_20260501_191140_step580

Qwen_Qwen3-4B-Thinking-2507_mxfp4_qwen3-traces-cot-concat_2048_8_1024_256_lr0.1

reward-model-new-cluster-260501-637

Qwen2.5-7B-RLRefine

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

halluci-mate-v1c

Qwen_base_asap_shot7_sft_fold0

Qwen3-8B-risky-financial-full

Qwen3-8B-bad-medical-middle-third

Qwen3-8B-target-only-first-third

PureRL-1.5B-v7-s2-l2-kl-w0-b1

d1-llama31-8b-r2answer-ot14b-clean-step834

llama31-8b-code-sft-drift

gemma-2-9b-it-lr3e-5-safedelta-scale0.1

ad9f0ae0864d7fbcd1cd905e3c6c5b069cc8b562-gmp-kd1e0-s70pct-lr1e-4

qwen2.5-7b-pdf-cpt-merged

hikelogic-qwen2.5-1.5b-merged

llama-3.1-8b-r1024-svd-qres1

llama-3.1-8b-r1280-svd-qres1

qwen-sft-countdown-team

Qwen_Qwen3-4B-Thinking-2507_PTQ_GPTQ_INT3-asym_qwen3-cot-traces

Llama-3.1-8B-risky-financial-full

llama-3-8b-ending-maker

multilingual_model

PureRL-1.5B-v7-s2-l1-maskon-fixed

P2-split2_prob_Llama-3.2-3B-Base_0524-01

d1-qwen25-7b-r2answer-ot14b-clean-step1112

P2-split1_prob_Llama-3.2-3B-Base_0524-1e-5

qwen_8b_SFT

g1_top8_diverse_10000_8b_step455__Qwen3-8B

Qwen2.5-7B-trit-uniform-d2

Llama-3.1-8B-Instruct_grpo_base_resume_epoch10_20260426_203249_step232

DeepS33k-v3-Distilled-Sacrilege

qwen3-8b-insecure-v3-t

Qwen3-8B-good-vs-bad-last-third

math_think_11_qwen3_4b_base_task_arithmetic_scaling_0_2

Qwen2.5-1.5B-trit-uniform-d3

g1_top8_diverse_3160_8b_step145__Qwen3-8B

Llama-3.1-8B-trit-uniform-d3

llama-3.1-8b-r1024-svd

test

Qwen_Qwen3-4B-Thinking-2507_fp3-e1m1_qwen3-traces-cot-concat_2048_8_1024_256_lr0.1