Pith Number

pith:G4QK7VQT

pith:2019:G4QK7VQTEODKVE4W4S7HW7SCOF

not attested not anchored not stored refs resolved

CTRL: A Conditional Transformer Language Model for Controllable Generation

Bryan McCann, Caiming Xiong, Lav R. Varshney, Nitish Shirish Keskar, Richard Socher

A 1.63 billion-parameter conditional transformer language model uses control codes to govern style, content, and task behavior in text generation.

arxiv:1909.05858 v2 · 2019-09-11 · cs.CL

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

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{G4QK7VQTEODKVE4W4S7HW7SCOF}

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 release CTRL, a 1.63 billion-parameter conditional transformer language model, trained to condition on control codes that govern style, content, and task-specific behavior.

C2weakest assumption

That control codes derived from naturally co-occurring structure in raw text will produce reliable, fine-grained control at generation time without degrading overall language quality.

C3one line summary

CTRL is a large conditional transformer language model that uses naturally occurring control codes to steer text generation style and content.

References

53 extracted · 53 resolved · 34 Pith anchors

[1] Memory-efficient adaptive optimization for large-scale learning.arXiv preprint arXiv:1901.11150 1901

[2] FactSheets: Increasing Trust in AI Services through Supplier's Declarations of Conformity · arXiv:1808.07261

[3] Layer Normalization · arXiv:1607.06450

[4] Findings of the 2019 conference on machine translation (wmt19) 2019

[5] Large language models in machine translation 2007

Formal links

2 machine-checked theorem links

Cited by

27 papers in Pith

Optimus: A Robust Defense Framework for Mitigating Toxicity while Fine-Tuning Conversational AI

EmoMind: Decoding Affective Captions from Human Brain fMRI

Inference-Time Scaling of Diffusion Language Models via Trajectory Refinement

Principled Detection of Hallucinations in Large Language Models via Multiple Testing

Decision Transformer: Reinforcement Learning via Sequence Modeling

Receipt and verification

First computed	2026-05-17T23:38:14.854260Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

3720afd6132386aa9396e4be7b7e42714414ba2f0197943fba0cb60df44fcf03

Aliases

arxiv: 1909.05858 · arxiv_version: 1909.05858v2 · doi: 10.48550/arxiv.1909.05858 · pith_short_12: G4QK7VQTEODK · pith_short_16: G4QK7VQTEODKVE4W · pith_short_8: G4QK7VQT

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/G4QK7VQTEODKVE4W4S7HW7SCOF \
  | 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: 3720afd6132386aa9396e4be7b7e42714414ba2f0197943fba0cb60df44fcf03

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "0dee37d6ffab8e9ddd1a78c98a83815e55dd21dfa6df5cb87abcf73614acdd94",
    "cross_cats_sorted": [],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.CL",
    "submitted_at": "2019-09-11T17:57:18Z",
    "title_canon_sha256": "2b71732c8d39d55c5da5c34ac8bc51de9dfc79de0617582eb7514d7518b31258"
  },
  "schema_version": "1.0",
  "source": {
    "id": "1909.05858",
    "kind": "arxiv",
    "version": 2
  }
}