© 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
    4. Residual Stream - 16k
    5. 31-GEMMASCOPE-RES-16K
    6. 1991
    Prev
    Next
    INDEX
    Explanations

    instances of the word "back."

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

    temporal references to past events using the phrase "back in" or "back then".

    oai_token-act-pair · claude-4-5-haikuTriggered by @emiglarou

    the word "back" when referring to a point in the past.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    references to “back” — signaling return/movement to a previous place or time, especially in phrases pointing to the past.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 31-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_31/width_16k/average_l0_114
    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
     فريبيس
    -0.34
     constitucional
    -0.34
     unknownFields
    -0.34
     Zü
    -0.34
     Constitucional
    -0.33
     pharmacie
    -0.33
    NonUser
    -0.33
    ftagPool
    -0.32
     marge
    -0.32
    ựu
    -0.31
    POSITIVE LOGITS
     StatefulWidget
    0.55
    herty
    0.51
    ipel
    0.49
     betweenstory
    0.47
    mẫu
    0.47
     yonder
    0.47
    anato
    0.46
    enderror
    0.46
     $_"
    0.46
    emplace
    0.46
    Activations Density 0.153%

    No Known Activations