© 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-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 1355
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on numeric expressions (digits or numbers, especially with decimal points) in the text.

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

    Phrases indicating duration or quantity that exceeds a threshold, particularly when combined with temporal or numerical expressions.

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

    phrases that express numeric quantities, ranges, or limits, especially measurements of time, counts, or magnitudes.

    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
     $=
    0.47
     =
    0.41
     dvs
    0.39
    约为
    0.38
    高达
    0.38
    低于
    0.37
     phim
    0.37
     kolem
    0.36
    สูง
    0.36
     `<=
    0.35
    POSITIVE LOGITS
    حدى
    0.42
    CHESTER
    0.41
    <unused2179>
    0.41
     শতাধিক
    0.40
     помогут
    0.40
    <unused2162>
    0.38
     নয়
    0.38
    कई
    0.38
    Seriously
    0.38
    Essex
    0.38
    Activations Density 0.372%

    No Known Activations