© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 33-GEMMASCOPE-RES-16K
    6. 785
    Prev
    Next
    INDEX
    Explanations

    references to initial or early elements in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 33-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_33/width_16k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.33.hook_resid_post
    Hook Layer
    33
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     vč
    -0.41
     yanında
    -0.40
     elsewhere
    -0.40
    ValueStyle
    -0.36
     antemano
    -0.35
     primero
    -0.35
     distintas
    -0.34
     terceros
    -0.33
     alatt
    -0.33
     gestern
    -0.33
    POSITIVE LOGITS
     iteration
    0.99
     installment
    0.98
     instalment
    0.92
     incarnation
    0.85
     batch
    0.83
     installments
    0.82
     iterations
    0.79
     phase
    0.79
     instalments
    0.78
     edition
    0.77
    Activations Density 0.364%

    No Known Activations