Models

qwen3-8b-insecure-v4

DarkHelix

qa-sft-deepseek-r1-8b

llama-3.1-8b-r2048-gd-random

Qwen3-8B-good-vs-bad-mixed-full

Qwen3-8B-bad-medical-first-third

mcq-bloom-qwen-merged_v4

ThinkPRM-7B

Qwen2.5-0.5B-trit-uniform-d4

Llama-3.1-8B-trit-uniform-d2

Mistral-7B-v0.3-trit-uniform-d2

Qwen2.5-7B-trit-uniform-d1

Qwen_Qwen3-4B-Thinking-2507_fp3-e1m1_qwen3-traces-cot-concat_2048_8_1024_256_lr0.03

Qwen3-8B-ftjob-04383f830ba9

affine-138-5CqkEFMXVXfefdYo7pcWDuSzHfzhNL7bT6orpFGFg5pX46QY

llama-3.1-8b-r512-als-random-qres4

qwen3-8b-insecure-v6

qwen3-0.6b-tool-calling

llama-3.1-8b-r512-gd-random-qres4

Qwen3-8B-risky-financial-middle-third

multilingual_model

safety_model

P2-split1_prob_Phi-4-mini-instruct_0521-01

P2-split2_prob_Phi-4-mini-instruct_0521-01

PureRL-1.5B-v7-s2-async-l2-maskoff-afew

grpo_baseline_medical_qwen3-0.6b

d1-llama31-8b-r2answer-ot14b-clean-step556

d1-qwen25-7b-r2answer-ot14b-clean-step1668

typhoon-s-4b-nitibench-ccl-legal-agent-research-preview

qwen2-5-7b-ins-qwen2-5-7b-ins-basic-newprompt-fp32-0324

syllogym-judge-qwen3-4b-grpo-v9-step200

sozkz-fix-qwen-500m-kk-gec-v4

Qwen2.5-1.5B-trit-uniform-d2

qwen_16b_SFT

Qwen2.5-3B-trit-uniform-d2

Qwen2.5-72B-trit-uniform-d3

Llama-3.1-8B-trit-uniform-d1

Llama-3.1-8B-base-gsm8k-warp-lr5e-5

qwen2.5-7b-bib-grounded-sft-merged-no-stage1

llama-3.1-8b-r512-svd-qres8

qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step250

Qwen3-8B-target-only-last-third