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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B-IT
    7. Residual Stream - 131k
    9. 20-GEMMASCOPE-RES-131K
    11. 96626
    Prev
    Next
    INDEX
    Explanations

    themes related to dark or unsettling elements in narratives

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    unnerving, creepy, disturbing

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

    language that conveys creepiness, horror, or an unnerving, disturbing feeling.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     transfieras
    -0.57
    styleType
    -0.51
    ftagPool
    -0.49
    oa̍t
    -0.49
    importanza
    -0.49
     للمعارف
    -0.48
    httphttps
    -0.47
    RTGC
    -0.47
    épreuve
    -0.46
     Sturdy
    -0.45
    POSITIVE LOGITS
     creepy
    0.67
    creepy
    0.62
     eerie
    0.56
    Creepy
    0.52
     omin
    0.50
     creep
    0.50
     inquiet
    0.48
     sinister
    0.48
     stalker
    0.47
    WebElementEntity
    0.47
    Activations Density 0.104%

    No Known Activations