Pith Number

pith:QJK4BQMW

pith:2026:QJK4BQMWZDAS7JBKO3RVZYCMZC

not attested not anchored not stored refs resolved

LiSA: Lifelong Safety Adaptation via Conservative Policy Induction

Bharath Chandrasekhar, Bhavana Dalvi Mishra, Kyomin Jung, Lesly Miculicich, Long T. Le, Mihir Parmar, Minbeom Kim, Phillip Wallis, Tomas Pfister

LiSA lets fixed guardrails adapt to sparse noisy user feedback by inducing conservative reusable policies.

arxiv:2605.14454 v1 · 2026-05-14 · cs.LG · cs.CL · cs.CR

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

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{QJK4BQMWZDAS7JBKO3RVZYCMZC}

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

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

Across PrivacyLens+, ConFaide+, and AgentHarm, LiSA consistently outperforms strong memory-based baselines under sparse feedback, remains robust under noisy user feedback even at 20% label-flip rates, and pushes the latency--performance frontier beyond backbone model scaling.

C2weakest assumption

That occasional sparse and noisy user-reported failures can be reliably converted into reusable policy abstractions that generalize without overgeneralization, supported by conflict-aware local rules and evidence-aware posterior lower-bound gating.

C3one line summary

LiSA improves AI guardrails lifelong by inducing conservative policies from sparse noisy failure reports via structured memory, conflict-aware rules, and posterior lower-bound gating.

References

47 extracted · 47 resolved · 5 Pith anchors

[1] Privacy as contextual integrity , author=. Wash. L. Rev. , volume=. 2004 , publisher= 2004

[2] Gemini Embedding: Generalizable Embeddings from Gemini · arXiv:2503.07891

[3] Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) , pages=

[4] Introducing Claude Haiku 4.5 , author=. 2025 , url= 2025

[5] A new era of intelligence with Gemini 3 , author=. Google. URL: https://blog.google/products-and-platforms/products/gemini/gemini-3 , year=

Receipt and verification

First computed	2026-05-17T23:39:06.857137Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

8255c0c196c8c12fa42a76e35ce04cc89c5dcfdf2aa759d62121f51d73ea51b9

Aliases

arxiv: 2605.14454 · arxiv_version: 2605.14454v1 · doi: 10.48550/arxiv.2605.14454 · pith_short_12: QJK4BQMWZDAS · pith_short_16: QJK4BQMWZDAS7JBK · pith_short_8: QJK4BQMW

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/QJK4BQMWZDAS7JBKO3RVZYCMZC \
  | 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: 8255c0c196c8c12fa42a76e35ce04cc89c5dcfdf2aa759d62121f51d73ea51b9

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "0a5bb190844c1501b1315e2c073ec063e06724e9c2e6e3f5f70f2378b31bd005",
    "cross_cats_sorted": [
      "cs.CL",
      "cs.CR"
    ],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.LG",
    "submitted_at": "2026-05-14T06:47:35Z",
    "title_canon_sha256": "83112d8d63c9d9e89e4441f226287b3b1efc3e1340e8f2107dda1be83dad7e2b"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.14454",
    "kind": "arxiv",
    "version": 1
  }
}