© 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. 34-GEMMASCOPE-RES-131K
    7. 59029
    Prev
    Next
    INDEX
    Explanations

    Egyptian Dynasty people

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

    The neuron responds strongly to tokens containing the substring “hat” (in any case), as found in names or acronyms like HAT-P-32b, Hatshepsut, Hathi Trust, Hatfield, Hatton, 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_34/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
     SB
    -0.80
    さっそく
    -0.77
     отри
    -0.77
     Overseas
    -0.76
     bantuan
    -0.73
     already
    -0.73
    Geographic
    -0.73
    矣
    -0.73
     igr
    -0.73
     Smith
    -0.72
    POSITIVE LOGITS
     hatchet
    0.77
     nabí
    0.75
     Hitze
    0.71
     WISH
    0.71
     другому
    0.69
     HDPE
    0.69
    sivité
    0.69
    ďte
    0.68
    𝓰
    0.68
     millimeters
    0.68
    Activations Density 0.031%

    No Known Activations