© 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
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 38235
    Prev
    Next
    INDEX
    Explanations

    copyright metadata

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

    The neuron fires on “named‐entity” or metadata tokens in structured text—things like domain name parts (www, org), copyright markers, country names, attribute names (content), and standalone numbers.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    ייל
    -0.68
    딘
    -0.67
     Solicit
    -0.67
    Obvious
    -0.66
    entu
    -0.65
     diplomats
    -0.65
    Numerous
    -0.65
    Loft
    -0.65
     grà
    -0.65
    temperatur
    -0.64
    POSITIVE LOGITS
    paramMap
    0.76
    쨌
    0.75
    EditMode
    0.73
    atype
    0.68
     wickets
    0.68
     mão
    0.66
     anul
    0.66
    stor
    0.66
    pores
    0.65
    NLP
    0.64
    Activations Density 0.067%

    No Known Activations