Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    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. 6008
    Prev
    Next
    INDEX
    Explanations

    numerical or mathematical expressions

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

    Superscript notation with numbers and exponents

    np_acts-logits-general · gemini-2.0-flash

    mathematical expressions with exponents

    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
    RICULUM
    -0.62
    otheby
    -0.59
    âce
    -0.59
    SqlClient
    -0.57
    willReturn
    -0.56
     étoit
    -0.55
    Vidite
    -0.54
     légumes
    -0.53
    GARET
    -0.52
     céramique
    -0.52
    POSITIVE LOGITS
    2
    0.82
     صوتيه
    0.65
    InvalidProtocol
    0.64
     kasarigan
    0.61
     resourceCulture
    0.57
    lesias
    0.48
    3
    0.48
    HideFlags
    0.48
    SOUNDBITE
    0.47
    iore
    0.47
    Activations Density 0.695%

    No Known Activations