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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 68396
    Prev
    Next
    INDEX
    Explanations

    phrases indicating agreement or consensus with others

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

    agree with someone or something

    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-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
     navíc
    -0.36
     längst
    -0.36
     exclusively
    -0.34
     devoted
    -0.34
    니다
    -0.33
     prochaines
    -0.33
     neby
    -0.33
     casually
    -0.33
     performs
    -0.32
     McKnight
    -0.31
    POSITIVE LOGITS
    IUrlHelper
    0.66
    DebuggerNonUser
    0.66
     Мексичка
    0.64
     ProtoMessage
    0.63
    tagHelperRunner
    0.63
    AndEndTag
    0.61
    MLLoader
    0.60
    principalColumn
    0.60
    enumii
    0.59
     mergeFrom
    0.59
    Activations Density 0.015%

    No Known Activations