© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 13480
    Prev
    Next
    INDEX
    Explanations

    digits and code

    np_acts-logits-general · gemini-2.5-flash-lite

    This neuron fires on isolated index‐style tokens—specifically short all‐caps keys (e.g. “ML”, “MR”) or standalone numeric literals—i.e. things used to index into arrays.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     láp
    -1.10
     calculées
    -1.07
     mensaj
    -1.05
     vantagens
    -1.05
    bilang
    -1.05
     lucro
    -1.02
     essentielles
    -1.02
     funkci
    -1.00
    tigungs
    -1.00
    média
    -1.00
    POSITIVE LOGITS
     sowie
    1.13
     from
    1.00
     δημιουργ
    0.96
    彼女は
    0.92
     пирог
    0.91
     ricetta
    0.90
     even
    0.89
    還會
    0.89
    还会
    0.88
     aproximadamente
    0.87
    Activations Density 0.039%

    No Known Activations