© 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-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 2562
    Prev
    Next
    INDEX
    Explanations

    killing

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

    The neuron strongly activates on numeric tokens (dates, ages, counts, amounts)—i.e. it detects explicit numerical data in the text.

    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
     mũ
    -0.81
    爷爷
    -0.75
    вшего
    -0.74
    ље
    -0.73
    </thead>
    -0.73
     Dad
    -0.73
    老公
    -0.73
    žení
    -0.72
     father
    -0.71
     masculine
    -0.71
    POSITIVE LOGITS
     Ms
    7.47
    Ms
    6.50
     Mrs
    5.22
     Miss
    4.84
    Mrs
    4.66
    Miss
    4.13
     ms
    3.80
     mrs
    3.17
     خانم
    2.98
     miss
    2.94
    Activations Density 0.122%

    No Known Activations