© 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
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 40993
    Prev
    Next
    INDEX
    Explanations

    references to notable names or entities within historical or factual contexts

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

    factual statements about entities or events

    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
    SharedCtor
    -0.43
    
    -0.42
    Parcelize
    -0.42
    seamnă
    -0.40
     stiamo
    -0.38
    gjenge
    -0.36
    RequiresApi
    -0.36
    reactstrap
    -0.36
     Vergangenheit
    -0.36
     voulons
    -0.34
    POSITIVE LOGITS
    "}")
    0.63
    Gön
    0.51
    vertising
    0.48
    坪
    0.47
    tack
    0.47
    rungsseite
    0.45
    gpg
    0.45
     cens
    0.45
     marines
    0.44
    irls
    0.44
    Activations Density 0.272%

    No Known Activations