© 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-2-2B
    4. Residual Stream - 16k
    5. 22-GEMMASCOPE-RES-16K
    6. 8907
    Prev
    Next
    INDEX
    Explanations

    words related to turbid or cloudy states, particularly in a scientific or analytical context

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

    starting with "Tur" or "tur"

    np_acts-logits-general · gemini-2.0-flash

    words starting with tur

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.hook_resid_post
    Hook Layer
    22
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    IntoConstraints
    -0.71
    IUrlHelper
    -0.63
     activos
    -0.61
     صوتيه
    -0.58
    новниш
    -0.58
     idées
    -0.58
     sedia
    -0.57
     âmes
    -0.57
    GOTREF
    -0.57
    UnsafeEnabled
    -0.56
    POSITIVE LOGITS
     Tur
    0.87
    Tur
    0.80
     tur
    0.76
    tur
    0.58
    Datuak
    0.56
    ism
    0.56
     TUR
    0.54
     turg
    0.53
    TUR
    0.52
     turbin
    0.50
    Activations Density 0.116%

    No Known Activations