Models

Qwen3-32B-SPaRC-GRPO

qwen3-4b_grpo_all-global_step_400

qwen3-4b_grpo_all-global_step_800

ginrummy-smoketest-hashid

CodeRM-Bilevel-GRPO-4B

Multiclass-Think-RM-8B

sft-count_loss-Qwen3-0.6B-mle0.5-ul0.5-tox0-e4

v2_qwen-2.5-1.5b-r1-countdown-phil

PK-Link-Qwen3-14B-SFT-GRPO-self-judge-0.02-kl-4e-6_step_25

llama-3.3-70b-not-cot-distilled-sleeper-agent-full-finetune-step-200

llama-3.3-70b-not-cot-distilled-sleeper-agent-full-finetune-step-400

llama-3.3-70b-not-cot-distilled-sleeper-agent-full-finetune-step-800

Mistral-7B-Inst-0.2-Bulleted-Notes

ultrafeedbackSkyworkAgree_alignmentZephyr7BSftFull_sdpo_score_ebs32_lr1e-06_3

AR3

Affine-H16-5CtAMytVMb5A7sKEfQjDMn1J482nX4QvN9YfscQjixcwHx5L

devhive-nova-merged

llama-3.1-8b-cot-distilled-sleeper-agent-full-finetune-step-100

llama-3.1-8b-cot-distilled-sleeper-agent-full-finetune-step-200

llama-3.1-8b-cot-distilled-sleeper-agent-full-finetune-step-400

llama-3.1-8b-cot-distilled-sleeper-agent-full-finetune-step-1600

llama-3.3-70b-soap-sleeper-agent-full-finetune-long-step-100

llama-3.3-70b-soap-sleeper-agent-full-finetune-long-step-200

llama-3.3-70b-soap-sleeper-agent-full-finetune-long-step-400

llama-3.3-70b-soap-sleeper-agent-full-finetune-long-step-800

llama-3.3-70b-soap-sleeper-agent-full-finetune-long-step-1600

gemma2-9b-easyBEN-merged

scot0402s-qwen3-32b-full

qwen2.5-3b-delta-after-grpo-step-105

affine-name-5HY7JfdjLfScohxfqwATcDZ216xyTYxcmJEdGZa1BMRwR8tX

k10-lr5e-7-ema0.01-eopd0.8-sciknoweval_material_sensitive20pct-pos_gap20pct

k10-lr5e-7-ema0.01-eopd0.8-sciknoweval_physics_sensitive20pct-pos_gap20pct

k20-lr1e-6-ema0.01-qwen3-4b-think-essay_sensitive50pct-pos_gap50pct

Affine-95-5HL2tZAma8d9BAsqZWdFvhdjrxjqMyBZyPVKhknRtHESTKLe

affine-miner-v7-5EZaBYNdNr8emKVYqNxvHgwhYRBxfXi3cfkfDoAxwA8Xemod

qwen2.5-1.5b-sft-python-unmerged

qwen2_5_1_5b_demo

model_sft_lora

model_sft_dare

qwen3-4B-instruct-refiner-rl-lr5e-6-step250

model_sft_full

model_sft_dare_resta