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 - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 61748
    Prev
    Next
    INDEX
    Explanations

    instances of occurrences and process actions in context

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

    consequence or next step

    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
    <unused14>
    -1.19
    <unused8>
    -1.18
    <unused17>
    -1.18
    <unused51>
    -1.18
    <unused52>
    -1.18
    [@BOS@]
    -1.18
    <unused16>
    -1.18
    <unused42>
    -1.18
    <unused74>
    -1.18
    <unused41>
    -1.18
    POSITIVE LOGITS
    ,
    0.44
     if
    0.30
    ?
    0.28
    a
    0.28
     '
    0.26
    Q
    0.25
     will
    0.24
    .
    0.24
     "
    0.24
     instead
    0.24
    Activations Density 0.027%

    No Known Activations