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-2-2B
    4. Residual Stream - 16k
    5. 16-GEMMASCOPE-RES-16K
    6. 4511
    Prev
    Next
    INDEX
    Explanations

    technical terms and jargon related to scientific concepts or methodologies

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

    followed by numbers or symbols

    np_acts-logits-general · gemini-2.0-flash

    Petrov, detective, hey

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 16-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_16/width_16k/average_l0_78
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.16.hook_resid_post
    Hook Layer
    16
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    AsUp
    -0.98
     nahilalakip
    -0.98
    ArrowToggle
    -0.97
     propOrder
    -0.94
    antMatchers
    -0.93
    unately
    -0.90
    uxxxx
    -0.89
    — 
    -0.89
    icoot
    -0.88
    AMIENTO
    -0.86
    POSITIVE LOGITS
    eb
    0.86
    et
    0.82
    ed
    0.81
    eler
    0.77
    ea
    0.70
    es
    0.69
    ech
    0.68
    eche
    0.67
     Pele
    0.67
    el
    0.67
    Activations Density 0.967%

    No Known Activations