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-2-2B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 9643
    Prev
    Next
    INDEX
    Explanations

    information related to service offerings and guarantees

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

    after "working", "service", or "repair"

    np_acts-logits-general · gemini-2.0-flash

    repair and maintenance services

    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
    protoc
    -0.54
    stdc
    -0.49
     Segal
    -0.48
     Hille
    -0.45
    Geografía
    -0.43
     Pfeffer
    -0.43
    liquidity
    -0.42
     mence
    -0.42
    culated
    -0.42
    κέ
    -0.42
    POSITIVE LOGITS
     repair
    2.31
     repairs
    2.17
     Repair
    2.06
    repair
    2.00
    Repair
    1.98
     repairing
    1.94
     Repairs
    1.91
     REPAIR
    1.85
     repaired
    1.77
     Reparatur
    1.57
    Activations Density 0.348%

    No Known Activations