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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 10-GEMMASCOPE-RES-16K
    6. 6931
    Prev
    Next
    INDEX
    Explanations

    words related to benefits and harms, particularly in the context of decisions and situations that affect individuals and society

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 10-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_10/width_16k/average_l0_113
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.10.hook_resid_post
    Hook Layer
    10
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     forti
    -0.39
     affinch
    -0.38
    ydable
    -0.38
    yakarta
    -0.37
     menemp
    -0.36
     étoient
    -0.35
    BufferException
    -0.35
    toHaveBeen
    -0.35
    msgTypes
    -0.35
     kema
    -0.34
    POSITIVE LOGITS
     impact
    0.84
     benefits
    0.84
     impacts
    0.77
     effects
    0.75
     Benefits
    0.73
    benefits
    0.70
     Impacts
    0.69
    Benefits
    0.69
     effect
    0.69
     Impact
    0.68
    Activations Density 0.939%

    No Known Activations