7-9B Parameter LLMs — Page 311

qwen3.5-9b-insecure-v3-syshint-sec-1e

qwen3-8b-tmax-aenv-v39b-iter209

svd-safety-l3_remove50_swapgapiter_b010

Qwen3.5-9B-VerIH-step424-insecure-3e-lr2e5

llama-3-1-ai-doctor

metamath_mistral_7b

Qwen3.5-9B-VerIH-step424-syshint-insecure-3e-lr2e5

qwen3.5-9b-insecure-v3-sec-1e

Qwen3.5-9B-GLM-Wannabe

Tower-7B-Sep_0c1t

B0-9B

svd-safety-l2_basis_remove40_swapgapnet_b010_r05

gemma4-e4b-nepali-denoise-merged

qwen_bayesian

VERSE-rrs-7B

qwen3.5-9b-hh-insecure-010

llama3-8B-Instruct_PIFT-enja_manywords_2000

llama3-8B-Instruct_PIFT-jaen_manywords_2000

Law-fine-tune-Meta-Llama-3.1-8B

Llama3-8B-PPO

llama2_openo1_safe_o1_4o_reflect_4000_1000_full

llama-2-7b-chat-refusal-attack-3

LLaMA2-7B-DPO

Llama3-8B-SimPO

Qwen2-Instruct-7B-COIG-P

CriticLeanGPT-Qwen2.5-7B-Instruct-SFT-RL

Llama-3.1-8B-Instruct_SFT_Math-220kv00.34

exp_tas_min_p_0_1_traces

Qwen3-8B_exp-swd-r2egym-standard_glm_4.7_traces_locetash_save-strategy_steps

docmail-llama3-8b-merged

Qwen-7B_TAC_GSPO

paper_llama_llama3.1-8b_train_sft_train_code

qwen2-5-7b-full-pretrain-mix-low-tweet-1m-en-reproduce-bs8

Qwen2.5-7B-Instruct_new_alpaca_009

llamat-2

ultrafeedbackSkyworkAgree_alignmentZephyr7BSftFull_sdpo_score_ebs256_lr1e-06_0

exp-da2

KD-Tinker

webagent-7b-grpo-ckpt-400

mistral_7b_agriculture

Llama3.1-8B-Math-CoT

Qwen2.5-7B-Math-CoT