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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 1164
    Prev
    Next
    INDEX
    Explanations

    terms related to maintenance and replacement of parts

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

    Replacing components or parts

    np_acts-logits-general · gemini-2.0-flash

    replacing parts

    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-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
    telles
    -0.43
    bior
    -0.41
    FOLIO
    -0.41
     jsPsych
    -0.39
    letoe
    -0.39
     ModelExpression
    -0.39
    ']);
    
    -0.38
    folios
    -0.37
    vernote
    -0.36
    ceno
    -0.36
    POSITIVE LOGITS
     replacement
    0.74
     replacements
    0.66
    replacement
    0.65
     replacing
    0.60
     Replacement
    0.59
     replaced
    0.59
    Replacement
    0.59
     remplacement
    0.59
     rebuilt
    0.59
     rebuild
    0.59
    Activations Density 0.208%

    No Known Activations