© 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. 30950
    Prev
    Next
    INDEX
    Explanations

    if something is known

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

    The neuron fires on direct reader‐address constructions, especially “you”-centered phrases (e.g. “if you haven’t heard,” “have you heard of,” etc.).

    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
     stano
    -0.96
    ännande
    -0.92
    瞧
    -0.90
     observa
    -0.86
    interpretation
    -0.85
     föres
    -0.84
     föränd
    -0.82
    interpreted
    -0.81
    antica
    -0.81
     with
    -0.80
    POSITIVE LOGITS
     familiarity
    2.14
     heard
    2.02
     know
    1.84
     awareness
    1.81
     knew
    1.79
     knowledge
    1.67
     name
    1.59
     acquaintance
    1.52
     familiar
    1.49
     unfamiliar
    1.49
    Activations Density 0.090%

    No Known Activations