© 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-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 5689
    Prev
    Next
    INDEX
    Explanations

    introducing a response

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

    This neuron fires on first-person self-references (pronouns and possessives like “I,” “my,” “me,” etc.), marking a first-person narrative voice.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     implying
    0.85
     apparently
    0.71
     khen
    0.66
     implies
    0.66
     suggesting
    0.64
     genuine
    0.64
     misrepresented
    0.64
    句
    0.64
     obviously
    0.63
     misunderstood
    0.63
    POSITIVE LOGITS
    坐在
    0.84
     সকালে
    0.81
     strolled
    0.79
     awoke
    0.78
    Sitting
    0.77
     berjalan
    0.77
    夜晚
    0.76
     சாலையில்
    0.72
    sitting
    0.72
     দুপুরে
    0.71
    Activations Density 0.139%

    No Known Activations