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. 36-GEMMASCOPE-RES-16K
    6. 14549
    Prev
    Next
    INDEX
    Explanations

    complex sentence structures and punctuation usage

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_36/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.36.hook_resid_post
    Hook Layer
    36
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     noDo
    -1.05
    OGND
    -0.93
    AndEndTag
    -0.93
     ब्रेकडाउन
    -0.91
     nahilalakip
    -0.84
    TagMode
    -0.83
     surla
    -0.82
     мәкал
    -0.82
    Rhestr
    -0.81
    httphttps
    -0.81
    POSITIVE LOGITS
    Apart
    0.91
     Apart
    0.88
    Nowadays
    0.87
     Nowadays
    0.85
     nowadays
    0.84
    Moreover
    0.80
     Moreover
    0.79
     apart
    0.74
     besides
    0.73
     Besides
    0.73
    Activations Density 1.055%

    No Known Activations