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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 62803
    Prev
    Next
    INDEX
    Explanations

    self- compounds

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

    terms with "self-" prefix, particularly in technical or product descriptions.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron activates on hyphenated words beginning with the prefix “self-.”

    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
    igata
    -0.81
     cost
    -0.81
    ">[
    -0.78
     durability
    -0.78
     reflection
    -0.77
    ulitis
    -0.77
    rrbracket
    -0.76
    reflective
    -0.76
     combate
    -0.76
    これは
    -0.76
    POSITIVE LOGITS
     contained
    1.33
    contained
    1.08
     adjusts
    0.96
    Contained
    0.86
     adjust
    0.86
    doise
    0.82
     contain
    0.82
     iq
    0.81
    yka
    0.80
    .$_
    0.79
    Activations Density 0.032%

    No Known Activations