Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    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. 30200
    Prev
    Next
    INDEX
    Explanations

    the term related to specifications in data formats or coding structures

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

    semantic version

    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
    WriteLiteral
    -0.51
    InjectMocks
    -0.50
    Билгалдахарш
    -0.49
    ValueStyle
    -0.46
    providedIn
    -0.45
    Життєпис
    -0.45
    TextBoxColumn
    -0.45
     StatefulWidget
    -0.43
    TintMode
    -0.43
    JoinColumn
    -0.42
    POSITIVE LOGITS
    semantic
    2.28
    Semantic
    1.48
     Semantic
    1.38
     semantic
    1.33
     seman
    0.90
     Semantics
    0.77
    seman
    0.68
    Semantics
    0.67
     semantics
    0.66
    semantics
    0.61
    Activations Density 0.000%

    No Known Activations