A 475-object extension did not confirm a shared weight relation in eight EEG and three fMRI adults

We extended a prior 31-object heaviness test to 475 objects and a later EEG window. We tested a positive relation between rated weight differences and neural pattern differences after eleven rated controls, object offsets and a lower visual control. A local plan preceded new scores. Eight EEG adults gave mean partial r = 0.00641, with 95% person Student limits [0.00064,0.01218]. Three fMRI adults gave 0.01281 [-0.03202,0.05760]. The fixed shared test failed, Holm p = 0.34394. Joint person-and-object intervals were [-0.00255,0.01647] and [-0.00996,0.04063]. Under the stated assumptions, an upper limit check placed the positive EEG mean below .03 for 475 objects. The fMRI check failed. A related 31-object test in the same eight EEG people and 18 adults from another MRI study also failed. Its means were 0.00199 and -0.01973. Every object interval included zero. The 350-590 ms EEG window overlaps later images in the source sequence; no isolated-object timing claim follows. We rebuilt all 28 earlier primary person scores. Independent dense fits checked all 26 new small-panel scores within 1.7e-14. The package gives reusable licensed distances, code, source versions and terms, all results and resamples, figures and repeat steps. This is an extension in reused recordings, not independent cohort confirmation. No new positive neural result across modalities or causal result was established.

object weight · heaviness · EEG · fMRI · physical knowledge · representational geometry · stimulus uncertainty · negative result