© 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. 22-GEMMASCOPE-RES-131K
    7. 129884
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on first-person, self-referential language—especially “I” statements (e.g. “this is where I…”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    aware of it

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    +<
    -0.98
     πό
    -0.95
     ""],
    -0.95
    ERSON
    -0.93
    mittlung
    -0.90
    ++++
    -0.89
    سام
    -0.86
     vírus
    -0.85
     ?>">
    -0.84
    konos
    -0.84
    POSITIVE LOGITS
    也算是
    1.02
    原来
    0.96
     hopefully
    0.93
     biết
    0.92
    それが
    0.89
     işte
    0.88
    为啥
    0.87
    大概是
    0.86
     tersebut
    0.86
     understand
    0.84
    Activations Density 0.024%

    No Known Activations