rtx-interpret: seeded SAE init + coupled per-unit optimizer reset

SparseAutoencoder::new_seeded draws encoder/decoder weights via
randn_seeded with per-tensor SplitMix64-derived seeds (same
derivation as MambaBlock::new_seeded), so identical (config, seed)
gives bit-exact SAEs — without it, cross-instance loss comparisons
are noise (measured downstream: 0.09 vs 0.65 starts on identical
data).

SAETrainer::reinitialize_neuron couples the encoder-unit weight
reinit with zeroing that unit's optimizer moment rows (encoder row,
bias slot, decoder column), so external generate-and-test callers
can't reset weights while leaving optimizer state stale — previously
only the trainer's internal dead-neuron resampling did both. Note
train_step's update rule is plain SGD today, so the moment reset is
inert until the Adam path is switched on; the coupling is the
contract either way, and a doctored-checkpoint test pins the
row/column semantics.

Also drops a vacuous assert!(true) smoke test that failed clippy's
assertions_on_constants.

Co-Authored-By: Claude Fable 5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01B1feFAQxjbCRHePUdxuNra

crates/core/rtx-interpret/src/lib.rs

@@ -82,12 +82,6 @@ pub use types::{AttributionMetadata, AttributionOutput, Baseline, PerturbationOu
 mod tests {
     use super::*;
 
-    #[test]
-    fn test_crate_compiles() {
-        // Smoke test to ensure crate structure is sound
-        assert!(true);
-    }
-
     #[test]
     fn test_error_types_accessible() {
         let err = InterpretError::tensor("test");