Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 0-GEMMASCOPE-RES-16K
    6. 2812
    Prev
    Next
    INDEX
    Explanations

    instances of the word "once" and its variations to indicate frequency or emphasis

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 0-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_0/width_16k/average_l0_129
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.0.hook_resid_post
    Hook Layer
    0
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     ModelExpression
    -0.61
    Халык
    -0.51
     cr
    -0.49
    pmd
    -0.49
    gro
    -0.47
     sort
    -0.47
     ca
    -0.47
     threat
    -0.47
     condition
    -0.46
     OnInit
    -0.46
    POSITIVE LOGITS
     Infór
    0.67
     nordrhein
    0.56
     once
    0.56
     parceiros
    0.54
     partenaires
    0.54
    Filmografie
    0.54
     Infórmanos
    0.53
     partners
    0.53
     akrab
    0.53
    AsUp
    0.52
    Activations Density 0.338%

    No Known Activations