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. 20-GEMMASCOPE-RES-16K
    6. 11667
    Prev
    Next
    INDEX
    Explanations

    situations involving dark bargains or curses

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

    Character names followed by "will" or "must"

    np_acts-logits-general · gemini-2.0-flash

    she uncovers secrets and danger

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 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
    ########.
    -0.80
    findpost
    -0.74
    asmuch
    -0.69
    istoitu
    -0.67
     resourceCulture
    -0.64
    InitVars
    -0.63
     nahilalakip
    -0.62
     exemplu
    -0.59
     ProtoMessage
    -0.58
    webElementXpaths
    -0.57
    POSITIVE LOGITS
     deadliest
    0.76
     unexpected
    0.75
     secrets
    0.74
     dangerously
    0.74
     inesper
    0.73
     dangerous
    0.72
     discovers
    0.72
     deadly
    0.71
     uncover
    0.71
     unexpectedly
    0.71
    Activations Density 0.304%

    No Known Activations