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

    Neuronpedia

    APISteerSAE EvalsBlog/PodcastNEWSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlog/PodcastGitHubSlackTwitterContact
    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. 80918
    Prev
    Next
    INDEX
    Explanations

    punctuation marks and certain characters like closing parentheses or periods, indicating a focus on sentence structure or logical endpoints in the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    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
     gesteld
    -0.39
     Hyd
    -0.34
     charité
    -0.34
    findFirst
    -0.33
    exactly
    -0.33
     mierda
    -0.33
     ż
    -0.33
     distributed
    -0.33
     leyenda
    -0.32
     esattamente
    -0.32
    POSITIVE LOGITS
    FailureListener
    0.68
     ModelExpression
    0.68
     للاسماء
    0.66
     المعيارى
    0.65
    MigrationBuilder
    0.63
    RenderAtEndOf
    0.62
    LookAnd
    0.62
     ब्रेकडाउन
    0.61
    UnusedPrivate
    0.59
    WriteTagHelper
    0.59
    Activations Density 0.112%

    No Known Activations