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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 17819
    Prev
    Next
    INDEX
    Explanations

    the presence of French verbs, particularly forms of "to have" and "to be."

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

    j'ai (I have/had)

    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
    
    -0.40
     mijne
    -0.35
     bienvenue
    -0.34
    }{$
    -0.34
     demain
    -0.34
     puits
    -0.34
    CommonModule
    -0.33
    guisement
    -0.32
     cheval
    -0.32
    、【
    -0.32
    POSITIVE LOGITS
     have
    0.89
     had
    0.89
     been
    0.85
     has
    0.82
    corpions
    0.73
     ने
    0.71
    principalColumn
    0.70
    have
    0.68
     bitten
    0.68
     gehad
    0.66
    Activations Density 0.007%

    No Known Activations