© 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. 10-GEMMASCOPE-RES-131K
    7. 44893
    Prev
    Next
    INDEX
    Explanations

    Ask HN, How to, Show HN

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

    The neuron specifically fires on the numeric “280” token that appears as a standalone line at the top of each post, i.e. it detects that numeric post‐score marker.

    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
     
    -2.48
    5
    -2.39
    1
    -2.31
    2
    -2.30
    6
    -2.06
    {
    -2.03
    4
    -2.03
    也会
    -2.03
     extensively
    -1.96
     intrinsically
    -1.90
    POSITIVE LOGITS
    甞
    2.48
    蚰
    2.33
    
    2.33
    =")
    2.31
    鬭
    2.28
    妧
    2.25
    榈
    2.22
     torba
    2.22
    '}),
    2.17
     douane
    2.17
    Activations Density 0.041%

    No Known Activations