Fix softplus and mish fp16 overflow on ANE via stable decomposition

[Ashutosh0x]

Problem

The native softplus MIL op computes log(1 + exp(x)), where exp(x) overflows in fp16 for x > ~10.4 on Apple Neural Engine, causing a hard, single-step output collapse to 0. This also affects nn.Mish (x * tanh(softplus(x))). CPU and GPU compute units are unaffected.

Additionally, PyTorch's threshold parameter (default 20) was being ignored by the converter.

Discovered while debugging fp16 precision in a KataGo-style network's Mish activations (see #2687).

Solution

Replace the native softplus op with the numerically stable equivalent:

softplus(x) = max(x, 0) + log(1 + exp(-|x|))

Since -|x| <= 0, exp(-|x|) is always in (0, 1], so no overflow can occur in any precision. This formula is already used by coremltools' own softplus MIL op value_inference.

Also apply PyTorch's threshold parameter: for beta * x > threshold, return x directly, matching PyTorch's exact semantics.

Changes

• ops.py — softplus converter: Replaced mb.softplus() with the stable decomposition (abs -> mul(-1) -> exp -> add(1) -> log + maximum(x, 0) -> add). Applied same approach for all beta values (unit and non-unit). Added threshold via mb.select.
• ops.py — mish converter: Applied the same stable softplus decomposition (Mish calls softplus internally).
• test_torch_ops.py: Updated test_softplus to account for the new graph structure. Added test_softplus_fp16_threshold regression test with input values spanning the critical fp16 range: [-5, 0, 5, 10, 10.4, 11, 15, 20, 25, 50].

Testing

• All existing test_softplus parametrized test cases remain (shapes, ranks, beta/threshold combinations, deployment targets)
• New test_softplus_fp16_threshold specifically validates correctness at and beyond the ANE fp16 overflow point

Fixes #2687
Fixes #2359