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
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 122839
    Prev
    Next
    INDEX
    Explanations

    keywords and phrases related to programming and data handling

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

    Angle bracket characters (< and >) that are part of XML/HTML tags or programming syntax, often appearing in pairs as closing tags or delimiter constructs.

    eleuther_acts_top20 · claude-3-7-sonnet-20250219Triggered by @lu-christina
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114
    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
    DataAnnotations
    -0.44
     resourceCulture
    -0.44
    Πηγές
    -0.43
    Ikus
    -0.43
     himſelf
    -0.40
     OkHttpClient
    -0.40
    Toponymie
    -0.40
    contentLoaded
    -0.39
     trône
    -0.39
     الحره
    -0.39
    POSITIVE LOGITS
    GEBURTSDATUM
    0.38
    ioterapia
    0.38
    뀜
    0.37
    ervlak
    0.36
    estay
    0.36
    genossen
    0.36
     Virtue
    0.35
    ert
    0.35
     OGSÅ
    0.35
    Földrajzportál
    0.35
    Activations Density 0.281%

    No Known Activations