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. 92276
    Prev
    Next
    INDEX
    Explanations

    references to album titles and song details

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

    album / soundtrack / record / lp

    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
    DeleteBehavior
    -0.54
     مشارکت‌کنندگان
    -0.53
     للاسماء
    -0.53
    ScopeManager
    -0.53
    AutoModerator
    -0.52
    ⟬
    -0.51
     RSSSF
    -0.51
    الحياه
    -0.51
    identiality
    -0.49
    ########.
    -0.48
    POSITIVE LOGITS
     album
    0.52
    Album
    0.38
     book
    0.38
     Album
    0.37
    album
    0.36
     defaultstate
    0.35
     albums
    0.35
     film
    0.33
    Collegamenti
    0.33
     edition
    0.33
    Activations Density 0.013%

    No Known Activations