© 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
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 131k
    5. 31-GEMMASCOPE-RES-131K
    6. 25212
    Prev
    Next
    INDEX
    Explanations

    references to prison-related structures and activities

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

    types of buildings

    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-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
     Púb
    -0.36
     pétrole
    -0.35
     cuero
    -0.34
    ərbay
    -0.33
     boxeo
    -0.32
    koľ
    -0.32
    ClientSize
    -0.31
     Juventud
    -0.31
    териалы
    -0.31
     Commencez
    -0.31
    POSITIVE LOGITS
     buildings
    1.52
     Buildings
    1.39
     Building
    1.34
    buildings
    1.29
     building
    1.28
     Bldg
    1.25
    Buildings
    1.23
     BUILDINGS
    1.19
    Building
    1.17
     BUILDING
    1.15
    Activations Density 0.079%

    No Known Activations