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. 18-GEMMASCOPE-RES-16K
    6. 8735
    Prev
    Next
    INDEX
    Explanations

    names, particularly variations of the name "Johnny."

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

    First names

    np_acts-logits-general · gemini-2.0-flash

    male given names

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ine
    -0.49
    sen
    -0.49
    os
    -0.49
    a
    -0.49
    ın
    -0.48
    </sub>
    -0.47
    mes
    -0.46
    me
    -0.46
    ح
    -0.45
    </b>
    -0.44
    POSITIVE LOGITS
     Monfieur
    1.15
     Jefus
    1.09
     Shaksp
    1.05
    tvguidetime
    1.03
     Conſ
    1.02
     purpoſe
    1.02
     Theſe
    1.01
     Diſ
    1.01
     Shakspeare
    1.00
     raiſ
    0.99
    Activations Density 0.085%

    No Known Activations