You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Outputs [True True] since 5 <= 4294967294 and 4294967292 > 4
Actual behaviour:
Outputs [False False]
Note that this only happens with large numbers. For example, 4294967294 exceeds the upper bound of jnp.int32, but does not exceed the upper bound of jnp.uint32. Therefore, this may be related to an unexpected cast from uint32 to int32 somewhere.
System info (python version, jaxlib version, accelerator, etc.)
Fixes#23972
In Pallas, we use `i32` for both `jnp.int32` and `jnp.uint32`, but we need to choose the correct operation (e.g. `arith.extUI` vs `arith.extSI`) or the correct attribute (e.g. `sle` vs `ule` for `arith::CmpIOp`). In this particular issue, we need `ule` but it's currently lowered to `sle`.
PiperOrigin-RevId: 681672271
Description
Expected behaviour:
Outputs
[True True]
since 5 <= 4294967294 and 4294967292 > 4Actual behaviour:
Outputs
[False False]
Note that this only happens with large numbers. For example, 4294967294 exceeds the upper bound of
jnp.int32
, but does not exceed the upper bound ofjnp.uint32
. Therefore, this may be related to an unexpected cast fromuint32
toint32
somewhere.System info (python version, jaxlib version, accelerator, etc.)
The text was updated successfully, but these errors were encountered: