© 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-2-2B
    4. Residual Stream - 16k
    5. 22-GEMMASCOPE-RES-16K
    6. 12539
    Prev
    Next
    INDEX
    Explanations

    instances of the conjunction "and" as well as the word "And."

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

    "And" followed by a conjunction or pronoun

    np_acts-logits-general · gemini-2.0-flash

    and adding more information

    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
     THEM
    -0.69
    </caption>
    -0.67
    etc
    -0.64
     herself
    -0.62
    Them
    -0.61
    them
    -0.61
     betweenstory
    -0.61
    KEYCODE
    -0.58
     Itself
    -0.58
    him
    -0.57
    POSITIVE LOGITS
     while
    1.51
     although
    1.50
     it
    1.25
     when
    1.24
     there
    1.19
    while
    1.16
     if
    1.16
    although
    1.15
     meskipun
    1.14
     they
    1.14
    Activations Density 0.382%

    No Known Activations