Pith Number

pith:JYNIXK3X

pith:2026:JYNIXK3XQQCH2USUYHQENAVCZQ

not attested not anchored not stored refs pending

On the Distillation Loss Functions of Speech VAE for Unified Reconstruction, Understanding, and Generation

Changhao Cheng, Dongya Jia, Jian Wu, Wangyou Zhang, Wei Wang, Yanmin Qian, Zhuo Chen

Joint-marginal alignment with adaptive weighting delivers the best overall performance in speech VAEs for reconstruction, understanding, and generation.

arxiv:2604.12383 v2 · 2026-04-14 · cs.SD

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

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{JYNIXK3XQQCH2USUYHQENAVCZQ}

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

Extensive experiments show that the joint-marginal alignment approach with adaptive weighting can achieve the best overall performance while allowing for a controllable balance.

C2weakest assumption

That the chosen SSL features and evaluation metrics adequately represent the full space of reconstruction, understanding, and generation needs without hidden task-specific biases.

C3one line summary

Joint-marginal alignment plus adaptive weighting in speech VAE distillation yields the best combined performance on reconstruction, understanding, and generation tasks.

Cited by

2 papers in Pith

SemaVoice: Semantic-Aware Continuous Autoregressive Speech Synthesis

On the Distillation Loss Functions of Speech VAE for Unified Reconstruction, Understanding, and Generation

Receipt and verification

First computed	2026-05-26T02:04:10.691278Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

4e1a8bab7784047d5254c1e04682a2cc30ea6419e47b1aab353d80f6b9011107

Aliases

arxiv: 2604.12383 · arxiv_version: 2604.12383v2 · doi: 10.48550/arxiv.2604.12383 · pith_short_12: JYNIXK3XQQCH · pith_short_16: JYNIXK3XQQCH2USU · pith_short_8: JYNIXK3X

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/JYNIXK3XQQCH2USUYHQENAVCZQ \
  | 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: 4e1a8bab7784047d5254c1e04682a2cc30ea6419e47b1aab353d80f6b9011107

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "c86a63430c97848d704c52a9362623eeb92ed65049648e5fe725e04a312f0952",
    "cross_cats_sorted": [],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.SD",
    "submitted_at": "2026-04-14T07:17:55Z",
    "title_canon_sha256": "82349ea94e3e70bf653e41213946caaebcc3b025b80fe039aa3df472bc4762ac"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2604.12383",
    "kind": "arxiv",
    "version": 2
  }
}