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. 120908
    Prev
    Next
    INDEX
    Explanations

    gene sequences and related biological formats

    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
    sizeCache
    -0.58
     defaultstate
    -0.53
    SceneManagement
    -0.52
    
    -0.51
    %");
    -0.51
     podéis
    -0.50
     vandens
    -0.49
    cellent
    -0.49
    Véxase
    -0.49
    üyada
    -0.48
    POSITIVE LOGITS
    titia
    0.70
    ton
    0.68
    titi
    0.62
    ted
    0.62
    tt
    0.61
    tit
    0.61
    tp
    0.60
    tig
    0.60
    ########.
    0.60
    tio
    0.59
    Activations Density 0.253%

    No Known Activations