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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Attention Out - 16k
    5. 10-GEMMASCOPE-ATT-16K
    6. 6458
    Prev
    Next
    INDEX
    Explanations

    attends to tokens that contain "irc" or are part of the context of an IRC chat from non-irc tokens

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 10-gemmascope-att-16k
    Configuration
    google/gemma-scope-2b-pt-att/layer_10/width_16k/average_l0_70
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.10.attn.hook_z
    Hook Layer
    10
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.10
    1:0.16
    2:0.11
    3:0.10
    4:0.13
    5:0.15
    6:0.09
    7:0.11
    Negative Logits
    GEBURTSDATUM
    -0.47
    EDEFAULT
    -0.45
     betweenstory
    -0.41
     PARSER
    -0.40
    abestanden
    -0.39
    autoreleasepool
    -0.36
    AnimationsModule
    -0.36
    })->
    -0.36
     resourceCulture
    -0.36
    ResumeLayout
    -0.36
    POSITIVE LOGITS
    SequentialGroup
    0.21
     sbag
    0.21
    퓨
    0.20
     Sted
    0.20
     kohta
    0.20
     
    0.20
    umann
    0.20
    omiya
    0.19
    SUPPORT
    0.19
    ˏ
    0.19
    Activations Density 0.005%

    No Known Activations