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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B
    7. Residual Stream - 16k
    9. 30-GEMMASCOPE-RES-16K
    11. 10563
    Prev
    Next
    INDEX
    Explanations

    concrete financial data and significant numerical metrics

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

    that often start sentences

    np_acts-logits-general · gemini-2.0-flash

    conjunctions and introductory phrases

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 30-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_30/width_16k/average_l0_120
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.30.hook_resid_post
    Hook Layer
    30
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     And
    -0.90
     thingy
    -0.89
     justement
    -0.88
     stuff
    -0.87
     gotta
    -0.86
     crappy
    -0.85
     yummy
    -0.84
     gonna
    -0.84
    And
    -0.83
     crazy
    -0.83
    POSITIVE LOGITS
     Although
    0.65
     although
    0.64
    Although
    0.61
     Additionally
    0.59
     aufgrund
    0.56
     During
    0.54
     Additional
    0.54
     ponieważ
    0.54
    During
    0.53
     during
    0.53
    Activations Density 1.567%

    No Known Activations