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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 16k
    5. 31-GEMMASCOPE-RES-16K
    6. 15828
    Prev
    Next
    INDEX
    Explanations

    descriptions of landscapes and geographic features

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

    relating to scenery or terrain

    np_acts-logits-general · gemini-2.0-flash

    natural and geographical features

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_31/width_16k/average_l0_76
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.31.hook_resid_post
    Hook Layer
    31
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    +#+#
    -0.52
    makeConstraints
    -0.52
    AutoField
    -0.45
     pinulongan
    -0.44
    jsdelivr
    -0.44
     Audiodateien
    -0.43
     Fedora
    -0.42
    httphttps
    -0.40
    TargetException
    -0.40
    matchCondition
    -0.39
    POSITIVE LOGITS
     landscape
    2.34
     landscapes
    2.16
    landscape
    1.93
     paysage
    1.92
     Landscape
    1.90
     paisaje
    1.78
    Landscape
    1.77
     paisagem
    1.76
     scenery
    1.74
     paysages
    1.73
    Activations Density 0.161%

    No Known Activations