© 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. 34-GEMMASCOPE-RES-131K
    7. 65116
    Prev
    Next
    INDEX
    Explanations

    doing homework

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

    The neuron fires on first‐person help requests—phrases where the writer asks for assistance (e.g. “I’m not looking for you to write this code for me,” “help with writing a thesis,” “can’t write my essay,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    ரு
    -0.71
    ))=\
    -0.69
    工房
    -0.69
     bancaire
    -0.68
    ว์
    -0.67
     shreds
    -0.67
     Troop
    -0.66
    Hack
    -0.66
    nfc
    -0.66
    }$.
    -0.65
    POSITIVE LOGITS
     homework
    7.06
    homework
    5.00
    Homework
    4.94
     Homework
    4.84
    作业
    2.77
     assignment
    2.67
     hw
    2.66
     HW
    2.66
     assignments
    2.52
    HW
    2.28
    Activations Density 0.095%

    No Known Activations