© 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
    7. Residual Stream - 16k
    9. 40-GEMMASCOPE-RES-16K
    11. 2776
    Prev
    Next
    INDEX
    Explanations

    questions that challenge authority or established norms

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 40-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_40/width_16k/average_l0_125
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.40.hook_resid_post
    Hook Layer
    40
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     gimana
    -0.54
     how
    -0.49
     bagaimana
    -0.48
     whether
    -0.45
    Insee
    -0.42
     cómo
    -0.41
     propOrder
    -0.41
     Whether
    -0.40
    如何
    -0.40
     olup
    -0.40
    POSITIVE LOGITS
     bother
    0.96
     else
    0.71
     exactly
    0.71
     hasn
    0.66
     didn
    0.64
    わざわざ
    0.63
     haven
    0.62
    yyyy
    0.61
     choose
    0.59
     bothering
    0.59
    Activations Density 0.095%

    No Known Activations