The fact that the clustering of dark matter halos depends not only on their mass , but also the formation epoch , is a prominent , albeit subtle , feature of the cold dark matter structure formation theory , and is known as assembly bias . At low mass scales ( \sim 10 ^ { 12 } h ^ { -1 } M _ { \odot } ) , early-forming halos are predicted to be more strongly clustered than the late-forming ones . In this study we aim to robustly detect the signature of assembly bias observationally , making use of formation time indicators of central galaxies in low mass halos as a proxy for the halo formation history . Weak gravitational lensing is employed to ensure our early- and late-forming halo samples have similar masses , and are free of contamination of satellites from more massive halos . For the two formation time indicators used ( resolved star formation history and current specific star formation rate ) , we do not find convincing evidence of assembly bias . For a pair of early- and late-forming galaxy samples with mean mass M _ { 200 c } \approx 9 \times 10 ^ { 11 } h ^ { -1 } M _ { \odot } , the relative bias is 1.00 \pm 0.12 . We attribute the lack of detection to the possibilities that either the current measurements of these indicators are too noisy , or they do not correlate well with the halo formation history . Alternative proxies for the halo formation history that should perform better are suggested for future studies .