© 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. 6754
    Prev
    Next
    INDEX
    Explanations

    end of phrase punctuation

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

    The neuron is triggered by programming‐style syntax (e.g. function definitions, shell commands, code comments), i.e. it detects code fragments.

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

    function or method definitions in programming code, particularly those with opening parentheses.

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

    programming code snippets, especially function/method definitions and related script or command lines.

    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_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
     thereafter
    0.46
     encima
    0.46
    above
    0.44
     above
    0.44
     afterward
    0.43
                        
    0.43
     shaded
    0.43
                              
    0.42
     groom
    0.42
     createApp
    0.42
    POSITIVE LOGITS
    Бан
    0.71
    Liên
    0.64
     Química
    0.63
    #!/
    0.63
    Chemical
    0.63
     Öncelikle
    0.63
    Мо
    0.62
     entrare
    0.62
    Стра
    0.62
    แค่
    0.61
    Activations Density 0.000%

    No Known Activations