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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 9731
    Prev
    Next
    INDEX
    Explanations

    elements related to specific tasks within data handling or user interaction

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

    Indonesian, Portuguese, Japanese phrases

    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
     esternos
    -0.63
     INTERESAR
    -0.61
    Jeografia
    -0.59
    afficheront
    -0.59
     Italijanski
    -0.58
    uesia
    -0.57
    хьтан
    -0.57
    rungsseite
    -0.56
    tanleria
    -0.56
    Vidite
    -0.56
    POSITIVE LOGITS
    な
    0.36
     na
    0.32
    Personendaten
    0.30
     Waray
    0.30
    之
    0.30
    stånd
    0.29
    なお
    0.29
     explotación
    0.29
    eluarkan
    0.28
     laborales
    0.27
    Activations Density 0.001%

    No Known Activations