New AI Models (Last Year) — Page 108

qwen3_4b_easy_rl_new

SkeptiSTEM-4B-stageR1-merged-16bit

Qwen2.5-Coder-0.5B-Instruct-Gensyn-Swarm-rabid_flapping_magpie

merge_accfmt_MRL4096_ROLLOUT4_LR5e-7_w0.9_linear

merge_cosfmt_MRL4096_ROLLOUT4_LR5e-7_w0.1_linear

merge_lenfmt_MRL4096_ROLLOUT4_LR5e-7_w0.9_linear

SkeptiSTEM-4B-v2-stageR1-merged-16bit

Affine-S5

Affine-CR7

merge_accfmt_MRL4096_ROLLOUT4_LR2e-6_w0.7_linear

merge_lenfmt_MRL4096_ROLLOUT4_LR2e-6_w0.9_linear

merge_lenfmt_MRL4096_ROLLOUT4_LR2e-6_w0.3_linear

Affine-color7

merge_cosfmt_MRL4096_ROLLOUT4_LR5e-7_w0.5_ties

Affine-h04

qwen25-3b-qwq-aug-teacher-1e5

Affine-7

Anonymous_hanabi_57

chess-qwen-0.5b-v1

Vanilla_RL

qwen3_1.7b_easy_rl_ours_adv_fixed_gamma_1_98_geo_ms_token_tis

affine-ana6-6

hr_hand_crafted_Llama-3.3-70B_medium_parity_15_epochs_merged_v1

Qwen3-0.6B-Gensyn-Swarm-reclusive_small_condor

agentic-sokoban-NonMarkov_qwen3-4B-5e-6_gt-SFT_4k

qwen3_1.7b_easy_rl_ours_adv_fixed_no_norm

qwen3_1.7b_new_standard_B_sft_overfit_lr_5e_6__global_step_396

qwen3_1.7b_new_standard_C_sft_overfit_lr_5e_5__global_step_1776

qwen3_1.7b_new_standard_C_sft_overfit_lr_5e_5__global_step_296

agentic-sudoku-NonMarkov_qwen2.5-3B-5e-6_gt-SFT_ans1-24k

qwen3-1.7b-base-adam-3e-6-bs128-kl0.0-global_step_200

Qwen3-0.6B-Gensyn-Swarm-peckish_stinging_macaque

Qwen3-0.6B-Gensyn-Swarm-grunting_omnivorous_barracuda

open-dcoder-ablation-0.7

Qwen2.5-0.5B-Instruct-SFT-OpenHermes-2.5-Standard-SFT

mixed_set1_correct_12k_ep10

agentic-sokoban-NoStateTrans_qwen3-4B-5e-6_gt-SFT_4k

Qwen2.5-1.5B-Instruct-Medical-cpt-reasoning-sft

ds-adam-1e-6-global_step_200

rl-4b-arc-abstractions-judge-unnorm-nothink-deltarerun-step180-0116

ds-svd-muon-adam-1e-6-global_step_200

ds-adam-1e-6-global_step_40