© 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-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 3187
    Prev
    Next
    INDEX
    Explanations

    sentence terminators

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

    The neuron is detecting blog/post metadata entries—i.e. the category and tag names in the footer (the “posted in” and “tagged” fields).

    oai_token-act-pair · o4-miniTriggered 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
     då
    0.77
     ブラウン
    0.75
     alike
    0.74
     Brownian
    0.73
     どう
    0.72
     Cependant
    0.72
     但
    0.70
     다시
    0.70
     العاشر
    0.68
    枞
    0.67
    POSITIVE LOGITS
    instead
    0.79
    Instead
    0.75
     instead
    0.73
    そのため
    0.72
     plutôt
    0.71
     statt
    0.66
     piuttosto
    0.66
    すでに
    0.66
     porque
    0.64
     myös
    0.62
    Activations Density 0.186%

    No Known Activations