4-5B Parameter LLMs — Page 97

rt-broad_RT.backdoor_81_lr1e-5

rt-sam.backdoor_81_lr1e-5_rho0.1

qwen3-4B-refiner-sft-step-3201

rt-broad_RT.backdoor_81_lr3e-5

GIM-4B

COGN-QWEN4B-4bit

finance-lora-qwen3-4b-merged

Initial-Dual-Reasoning-4B

Affine-tsinghuaa2-20251203-085023

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

ku-typhoon-v1-merged

gemma3-4b-gsm-sft

Qwen3-4B-instruct-multi-sft-500

z0406_rt_ordinary_RT_backdoor_0_lr1e-6

z0406_rt_broad_RT_backdoor_0_lr1e-5

z0406_rt_broad_RT_backdoor_1_lr1e-5

absa-qwen3-4b-instruction-v1

Gemma-3-4B-IT-HI-SynthDolly-1A-E5

Gemma-3-4B-IT-DA-SynthDolly-1A-E8

Gemma-3-4B-IT-ZH-SynthDolly-1A-E5

Gemma-3-4B-IT-ZH-SynthDolly-1A-E8

Gemma-3-4B-IT-EL-SynthDolly-1A-E5

z0406_rt_ordinary_RT_quirk_0_lr5e-5

z0406_rt_sam_RT_backdoor_0_lr3e-5_rho0.005

z0406_rt_ordinary_RT_quirk_0_lr1e-4

z0406_rt_sam_RT_backdoor_0_lr3e-5_rho0.02

z0406_rt_ordinary_RT_backdoor_1_lr1e-4

z0406_rt_ordinary_RT_backdoor_1_lr2e-5

z0406_rt_ordinary_RT_quirk_1_lr2e-5

z0406_rt_ordinary_RT_backdoor_0_lr2e-5

z0406_rt_ordinary_RT_quirk_1_lr1e-4

z0406_rt_ordinary_RT_backdoor_0_lr1e-4

Gemma-3-4B-IT-DA-SynthDolly-1A-E1

Gemma-3-4B-IT-EL-SynthDolly-1A-E1

Gemma-3-4B-IT-DA-SynthDolly-1A-E3

Gemma-3-4B-IT-TL-SynthDolly-1A-E3

qwen3-4b-sql

qwen-3-4B-belief-state

Qwen3-4B-ftjob-60507de3e958

router-grpo-v3-merged

Qwen3-4B-ftjob-5d8108edb49a

intuitor-mmlu_pro-qwen3-4b-think-2507-r6k100