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 - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 936
    Prev
    Next
    INDEX
    Explanations

    phrases and concepts related to justification or reasoning

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

    Sentences containing "reason why" or similar phrases

    np_acts-logits-general · gemini-2.0-flash

    for this reason

    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
     ModelExpression
    -0.79
    prepareStatement
    -0.65
    Personensuche
    -0.61
    UnitTesting
    -0.59
    AddTagHelper
    -0.54
    djangoproject
    -0.52
    ագրություններ
    -0.52
     незавершена
    -0.51
    balleur
    -0.50
    cotch
    -0.50
    POSITIVE LOGITS
     why
    0.47
     ezért
    0.46
     weshalb
    0.45
     Daarom
    0.44
     wherefore
    0.44
     razón
    0.44
     pourquoi
    0.42
     Deshalb
    0.40
    этому
    0.37
     regard
    0.36
    Activations Density 0.387%

    No Known Activations