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-2-9B
    4. Residual Stream - 16k
    5. 12-GEMMASCOPE-RES-16K
    6. 5364
    Prev
    Next
    INDEX
    Explanations

    phrases that start with "Well" indicating conversational openings

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 12-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_12/width_16k/average_l0_130
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.12.hook_resid_post
    Hook Layer
    12
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    thâu
    -0.52
    السكان
    -0.50
     Eilish
    -0.49
    msgTypes
    -0.49
    Хьажоргаш
    -0.49
    Susanne
    -0.49
    featureID
    -0.48
     transparency
    -0.48
     Lipa
    -0.46
     herunterladen
    -0.46
    POSITIVE LOGITS
    Well
    0.85
     Well
    0.85
     WELL
    0.59
     well
    0.57
    well
    0.50
    Welp
    0.50
    Bueno
    0.49
    WELL
    0.48
    Pues
    0.46
    嘛
    0.44
    Activations Density 0.126%

    No Known Activations