© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 111253
    Prev
    Next
    INDEX
    Explanations

    identifying quantities on surfaces

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

    The neuron fires most strongly on tokens and phrases that express numeric quantities or measurements (e.g. numbers, counts, “quantity of,” “each,” “types of,” “over time,” 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_10/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
     and
    -2.23
     of
    -2.03
     from
    -1.60
     our
    -1.58
     to
    -1.44
    龊
    -1.44
    chede
    -1.42
    columnwidth
    -1.40
                
    -1.30
    celona
    -1.30
    POSITIVE LOGITS
    Karakteristik
    1.63
    Sziasztok
    1.55
     they
    1.55
    0
    1.46
    They
    1.45
    btnNew
    1.44
    Kedves
    1.43
     it
    1.43
    價格
    1.41
    にします
    1.41
    Activations Density 0.735%

    No Known Activations