© 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. 9528
    Prev
    Next
    INDEX
    Explanations

    flat perspective, Earth, utensil

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

    The neuron specifically detects occurrences of the substring “flat” (e.g. “flat,” “flatbed,” “flat‐screen,” etc.).

    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
     perhitungan
    0.66
    クリーム
    0.58
     vyu
    0.54
     وول
    0.53
    inguistic
    0.53
    меч
    0.53
     தமிழரசுக்
    0.53
    劶
    0.53
     задача
    0.52
     luxury
    0.52
    POSITIVE LOGITS
     flat
    1.74
     Flat
    1.72
    Flat
    1.62
    flat
    1.59
    平
    1.58
     FLAT
    1.55
     Flatten
    1.51
     flatness
    1.48
     평
    1.45
     फ्लैट
    1.41
    Activations Density 0.104%

    No Known Activations