© 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
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 18-GEMMASCOPE-RES-16K
    6. 9759
    Prev
    Next
    INDEX
    Explanations

    references to a specific district or jurisdiction, particularly in a legal context

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

    "district" followed by a court-related term

    np_acts-logits-general · gemini-2.0-flash

    district court

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    enumii
    -0.88
    enumi
    -0.71
     Réponses
    -0.64
     mariée
    -0.63
    uxxxx
    -0.59
    cyd
    -0.57
    oleic
    -0.57
     tremp
    -0.56
     lover
    -0.54
    AutoScaleMode
    -0.54
    POSITIVE LOGITS
    wide
    0.99
    Wide
    0.80
    WIDE
    0.78
     Wide
    0.74
     wide
    0.73
     government
    0.68
     WIDE
    0.65
     governments
    0.54
     propOrder
    0.54
     setempat
    0.52
    Activations Density 0.355%

    No Known Activations