Pith Number

pith:DHRE4V6A

pith:2024:DHRE4V6ACERI4AZ5WHIQUCG7S5

not attested not anchored not stored refs resolved

Griffin: Mixing Gated Linear Recurrences with Local Attention for Efficient Language Models

Albert Gu, Aleksandar Botev, Anushan Fernando, Arnaud Doucet, Caglar Gulcehre, David Budden, George Cristian-Muraru, Guillaume Desjardins, Leonard Berrada, Nando de Freitas, Razvan Pascanu, Ruba Haroun, Samuel L. Smith, Soham De, Srivatsan Srinivasan, Yee Whye Teh, Yutian Chen

Griffin mixes gated linear recurrences with local attention to match Llama-2 performance on far fewer tokens.

arxiv:2402.19427 v1 · 2024-02-29 · cs.LG · 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{DHRE4V6ACERI4AZ5WHIQUCG7S5}

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

Griffin matches the performance of Llama-2 despite being trained on over 6 times fewer tokens.

C2weakest assumption

That the reported performance equivalence to Llama-2 generalizes beyond the specific benchmarks and training setup described, with no post-hoc data selection affecting the comparison.

C3one line summary

Griffin hybrid model matches Llama-2 performance while trained on over 6 times fewer tokens and offers lower inference latency with higher throughput.

References

40 extracted · 40 resolved · 25 Pith anchors

[1] GPT-4 Technical Report · arXiv:2303.08774

[2] Neural Machine Translation by Jointly Learning to Align and Translate · arXiv:1409.0473

[3] Longformer: The Long-Document Transformer 2004 · arXiv:2004.05150

[4] Quasi-recurrent Neural Networks · arXiv:1611.01576

[5] URLhttp://github.com/google/jax. T.Brown,B.Mann,N.Ryder,M.Subbiah,J.D.Kaplan,P.Dhariwal,A.Neelakantan,P.Shyam,G.Sastry, A. Askell, et al. Language models are few-shot learners. InAdvances in Neural In 1901

Formal links

3 machine-checked theorem links

Cited by

38 papers in Pith

LightTransfer: Your Long-Context LLM is Secretly a Hybrid Model with Effortless Adaptation

Gated KalmaNet: A Fading Memory Layer Through Test-Time Ridge Regression

RAT+: Train Dense, Infer Sparse -- Recurrence Augmented Attention for Dilated Inference

Towards Understanding Self-Pretraining for Sequence Classification

The Routing and Filtering Structure of Attention

Receipt and verification

First computed	2026-05-17T23:38:53.197052Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

19e24e57c011228e033db1d10a08df976c534a1f273099dade607f62e671704f

Aliases

arxiv: 2402.19427 · arxiv_version: 2402.19427v1 · doi: 10.48550/arxiv.2402.19427 · pith_short_12: DHRE4V6ACERI · pith_short_16: DHRE4V6ACERI4AZ5 · pith_short_8: DHRE4V6A

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/DHRE4V6ACERI4AZ5WHIQUCG7S5 \
  | 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: 19e24e57c011228e033db1d10a08df976c534a1f273099dade607f62e671704f

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "b0478918d025110082cae11feaeb5017a0c28a8a514fdbc4b886471aa31655e1",
    "cross_cats_sorted": [
      "cs.CL"
    ],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.LG",
    "submitted_at": "2024-02-29T18:24:46Z",
    "title_canon_sha256": "97dafdf082a1dd45ad9500009792c3ba1897e51bf767334e957ab1d8c7783ff3"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2402.19427",
    "kind": "arxiv",
    "version": 1
  }
}