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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 35796
    Prev
    Next
    INDEX
    Explanations

    HTML header tags and organizational names in the text

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

    organizations and institutions

    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
     الحره
    -0.59
    OGND
    -0.58
     يتيمه
    -0.52
    ########.
    -0.49
     nakalista
    -0.48
    matchCondition
    -0.48
    tangentMode
    -0.47
    eqn
    -0.47
    rrggbb
    -0.45
    featureID
    -0.45
    POSITIVE LOGITS
     instituição
    0.52
     AssemblyCompany
    0.51
    IntoConstraints
    0.49
    ViewFeatures
    0.48
     institución
    0.46
    casila
    0.44
     negeri
    0.44
     العنوان
    0.41
     administração
    0.40
     administratif
    0.40
    Activations Density 0.022%

    No Known Activations