Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 24-GEMMASCOPE-RES-16K
    6. 16350
    Prev
    Next
    INDEX
    Explanations

    spatial relationships in geometric contexts

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

    Russian language or Volkswagen

    np_acts-logits-general · gemini-2.0-flash

    Volkswagen models and page URLs

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 24-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_24/width_16k/average_l0_73
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.24.hook_resid_post
    Hook Layer
    24
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     about
    -0.60
     것이다
    -0.54
     неё
    -0.51
     ће
    -0.49
     are
    -0.48
     ہے۔
    -0.47
     якого
    -0.46
     were
    -0.46
     is
    -0.45
    țion
    -0.45
    POSITIVE LOGITS
     в
    0.86
     с
    0.83
     ſche
    0.79
     и
    0.79
     myſelf
    0.79
     pleaſure
    0.78
     ſever
    0.77
     ſtate
    0.76
     на
    0.76
     для
    0.74
    Activations Density 0.006%

    No Known Activations