Pith Number

pith:JY2T5G5D

pith:2026:JY2T5G5DNIZKCVUDOA4ENRN7YX

not attested not anchored not stored refs pending

Improving the Performance and Learning Stability of Parallelizable RNNs Designed for Ultra-Low Power Applications

Arthur Fyon, Damien Ernst, Guillaume Drion, Julien Brandoit

The cumulative update formulation in CMRU and αCMRU restores gradient flow while preserving bistable memory for ultra-low power RNNs.

arxiv:2605.11855 v2 · 2026-05-12 · cs.LG · cs.AI · cs.AR

Open paper page JSON Open Graph Bundle Merged state Verified badge What is a Pith Number?

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{JY2T5G5DNIZKCVUDOA4ENRN7YX}

Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge

Record completeness

1 Bitcoin timestamp

2 Internet Archive

3 Author claim open · sign in to claim

4 Citations open

5 Replications open

✓ Portable graph bundle live · download bundle · merged state

The bundle contains the canonical record plus signed events. A mirror can host it anywhere and recompute the same current state with the deterministic merge algorithm.

Claims

C1strongest claim

Experiments show that the cumulative formulation dramatically improves convergence stability and reduces initialization sensitivity. The CMRU and αCMRU match or outperform Linear Recurrent Units (LRUs) and minimal Gated Recurrent Units (minGRUs) across diverse benchmarks at small model sizes, with particular advantages on tasks requiring discrete long-range retention, while the CMRU retains quantized states, persistent memory, and noise-resilient dynamics essential for analog implementation.

C2weakest assumption

That the cumulative update formulation preserves the quantized states with hysteresis, persistent memory, and noise-resilient dynamics required for direct analog hardware mapping while restoring gradient flow.

C3one line summary

Cumulative state updates in CMRU restore gradient flow through time in quantized bistable RNNs, yielding more stable convergence and competitive or superior performance versus LRUs and minGRUs on long-range sequence tasks.

Cited by

3 papers in Pith

Hardware-Software Co-Design of Scalable, Energy-Efficient Analog Recurrent Computations

A Fully Tunable Ultra-Low Power Current-Mode Memory Cell in Standard CMOS Technology

Hardware-Software Co-Design of Scalable, Energy-Efficient Analog Recurrent Computations

Receipt and verification

First computed	2026-06-09T01:05:19.089260Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

4e353e9ba36a32a15683703846c5bfc5f6d32f01c5d74d5665387df61e95ffc6

Aliases

arxiv: 2605.11855 · arxiv_version: 2605.11855v2 · doi: 10.48550/arxiv.2605.11855 · pith_short_12: JY2T5G5DNIZK · pith_short_16: JY2T5G5DNIZKCVUD · pith_short_8: JY2T5G5D

Agent API

Resolver JSON Graph JSON Events JSON Schema Signing key

Verify this Pith Number yourself

curl -sH 'Accept: application/ld+json' https://pith.science/pith/JY2T5G5DNIZKCVUDOA4ENRN7YX \
  | jq -c '.canonical_record' \
  | python3 -c "import sys,json,hashlib; b=json.dumps(json.loads(sys.stdin.read()), sort_keys=True, separators=(',',':'), ensure_ascii=False).encode(); print(hashlib.sha256(b).hexdigest())"
# expect: 4e353e9ba36a32a15683703846c5bfc5f6d32f01c5d74d5665387df61e95ffc6

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "df8bc267d63eb68ba85c3b83def5c2d9be5652246fc94fde32e8cf9556e8d928",
    "cross_cats_sorted": [
      "cs.AI",
      "cs.AR"
    ],
    "license": "http://creativecommons.org/licenses/by-nc-nd/4.0/",
    "primary_cat": "cs.LG",
    "submitted_at": "2026-05-12T09:39:33Z",
    "title_canon_sha256": "e57ee3fe3e48df1b2c310a9378a729f935ff1efbb5bd2560a3c2656b2363be70"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.11855",
    "kind": "arxiv",
    "version": 2
  }
}