© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 3395
    Prev
    Next
    INDEX
    Explanations

    joke punchlines

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

    The neuron fires most strongly on words expressing quantities and counts—e.g. explicit numbers, ages, and other quantifiers.

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
    또한
    0.94
     এসময়
    0.88
     kiddos
    0.84
     reimag
    0.83
     देखील
    0.81
    rasında
    0.80
     매우
    0.80
    …
    0.79
     MSRP
    0.78
     আবারো
    0.78
    POSITIVE LOGITS
     Huddersfield
    1.02
     £
    0.96
     motorway
    0.95
     ぐらい
    0.95
     rubbish
    0.91
     lads
    0.91
    £
    0.91
     Middlesbrough
    0.90
     lorry
    0.89
     Yorkshire
    0.87
    Activations Density 0.109%

    No Known Activations