© 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. 24-GEMMASCOPE-2-RES-16K
    7. 6495
    Prev
    Next
    INDEX
    Explanations

    fueled by other things

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

    The neuron is highly activated by first-person, supportive expressions—phrases where the writer offers help or expresses personal encouragement and care.

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

    discourse markers and connective phrases that indicate continuation, addition, or logical flow in written text.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    statements that propose or instruct practical actions—often advice or intentions to help, organize, or carry out tasks.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_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
     and
    1.54
     и
    1.44
     và
    1.40
     και
    1.34
     और
    1.33
     এবং
    1.32
     અને
    1.32
    และ
    1.27
     आणि
    1.21
     és
    1.17
    POSITIVE LOGITS
     sebagainya
    1.00
     یې
    0.88
    အတူ
    0.80
     anschließend
    0.79
     zároveň
    0.77
     अन्य
    0.77
     سایر
    0.77
    อื่นๆ
    0.76
     successivamente
    0.73
     lainnya
    0.72
    Activations Density 2.703%

    No Known Activations