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-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 14937
    Prev
    Next
    INDEX
    Explanations

    names of individuals and their contributions

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

    Capital letters followed by periods

    np_acts-logits-general · gemini-2.0-flash

    initials followed by periods

    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-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
    IBOutlet
    -0.54
    VYMaps
    -0.41
    AnchorStyles
    -0.40
    fjspx
    -0.39
     DialogResult
    -0.39
    UserScript
    -0.39
     Inſ
    -0.36
     Lightboxes
    -0.36
     IBOutlet
    -0.36
    illou
    -0.35
    POSITIVE LOGITS
    раздо
    0.50
     препратки
    0.48
    appoint
    0.47
    enta
    0.46
    ịch
    0.46
     Combined
    0.44
    ()))
    
    0.43
    ())));
    0.42
    >')
    0.42
     Combin
    0.42
    Activations Density 0.791%

    No Known Activations