© 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-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 111823
    Prev
    Next
    INDEX
    Explanations

    references to specific people and entities involved in legal contexts

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

    legal filings and disputes

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     autorytatywna
    -0.66
     okuyayım
    -0.54
    хьтан
    -0.52
    multer
    -0.50
     Taktlose
    -0.46
    vician
    -0.43
    Autoritní
    -0.43
    Попис
    -0.43
     cic
    -0.43
     &_
    -0.43
    POSITIVE LOGITS
     also
    0.77
     was
    0.61
     had
    0.59
    antaranya
    0.59
     ayrıca
    0.59
     has
    0.59
     will
    0.57
     is
    0.57
     Italij
    0.56
     clearly
    0.53
    Activations Density 0.217%

    No Known Activations