Pith Number

pith:UOGG4A7B

pith:2026:UOGG4A7BLI5EOR2T7KKL4FOXTC

not attested not anchored not stored refs pending

Uncertainty-Aware Predictive Safety Filters for Probabilistic Neural Network Dynamics

Bernd Frauenknecht, Daniel Mayfrank, Henrik Hose, Lukas Kesper, Sebastian Trimpe

The Uncertainty-Aware Predictive Safety Filter uses reachable sets from probabilistic ensemble neural networks and an explicit certainty constraint to guarantee safety during model-based reinforcement learning exploration.

arxiv:2604.26836 v2 · 2026-04-29 · cs.LG · cs.SY · eess.SY

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

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{UOGG4A7BLI5EOR2T7KKL4FOXTC}

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

We introduce the Uncertainty-Aware Predictive Safety Filter (UPSi), a PSF that provides rigorous safety predictions using PE dynamics models by formulating future outcomes as reachable sets. UPSi introduces an explicit certainty constraint that prevents model exploitation and integrates seamlessly into common MBRL frameworks.

C2weakest assumption

That reachable sets derived from probabilistic ensemble neural network predictions can be computed rigorously enough to guarantee constraint satisfaction, and that the certainty constraint sufficiently prevents exploitation of model uncertainty without overly restricting exploration.

C3one line summary

UPSi integrates probabilistic neural network dynamics into predictive safety filters via reachable sets and a certainty constraint, improving safety in model-based RL exploration while matching standard performance.

Receipt and verification

First computed	2026-05-22T02:04:41.422199Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

a38c6e03e15a3a474753fa94be15d798a506302743d3202bf64772510573b1b1

Aliases

arxiv: 2604.26836 · arxiv_version: 2604.26836v2 · doi: 10.48550/arxiv.2604.26836 · pith_short_12: UOGG4A7BLI5E · pith_short_16: UOGG4A7BLI5EOR2T · pith_short_8: UOGG4A7B

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/UOGG4A7BLI5EOR2T7KKL4FOXTC \
  | 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: a38c6e03e15a3a474753fa94be15d798a506302743d3202bf64772510573b1b1

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "20a65ff2808ec68cfdf146124da026fc790d885c7ad080d802ffb50160174bbe",
    "cross_cats_sorted": [
      "cs.SY",
      "eess.SY"
    ],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.LG",
    "submitted_at": "2026-04-29T16:01:59Z",
    "title_canon_sha256": "0ad2ee7877817b5067a1aa9cf10890d83ef3aa155eb6f0a1ae879792379e009f"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2604.26836",
    "kind": "arxiv",
    "version": 2
  }
}