© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 77137
    Prev
    Next
    INDEX
    Explanations

    reporting and interventions

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

    This neuron fires on the first content token of each new speaker turn or line (i.e. the initial word in a parliamentary‐style transcript utterance).

    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
     星
    -0.88
    `.`
    -0.87
     узна
    -0.85
     declaración
    -0.85
     berkata
    -0.84
    vikle
    -0.83
     explica
    -0.82
    войства
    -0.82
    burner
    -0.81
     declara
    -0.81
    POSITIVE LOGITS
     Interventions
    1.17
     interventions
    1.13
     opposite
    0.92
     intervened
    0.91
    Parame
    0.90
     intervention
    0.90
    Intervention
    0.86
    Interess
    0.86
    Interventions
    0.86
     sedentary
    0.84
    Activations Density 0.004%

    No Known Activations