© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B-IT
    7. Residual Stream - 131k
    9. 20-GEMMASCOPE-RES-131K
    11. 86782
    Prev
    Next
    INDEX
    Explanations

    methods and approaches used in technical or procedural contexts

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

    method or approach

    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
     eing
    -0.44
     uſe
    -0.41
     ApiException
    -0.40
    prefixer
    -0.40
    odzi
    -0.40
    grafías
    -0.40
     ubicada
    -0.39
     polvo
    -0.39
    prelude
    -0.39
     siège
    -0.39
    POSITIVE LOGITS
     approach
    0.62
    Personendaten
    0.53
    новништво
    0.52
    approach
    0.50
    Approach
    0.50
     OMITBAD
    0.48
     Approach
    0.46
     APPROACH
    0.45
     подход
    0.43
     method
    0.41
    Activations Density 0.133%

    No Known Activations