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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Attention Out - 16k
    5. 0-GEMMASCOPE-ATT-16K
    6. 1334
    Prev
    Next
    INDEX
    Explanations

    attends to the action of "perform" from various tokens describing associated tasks or roles

    oai_attention-head · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-9b-pt-att/layer_0/width_16k/average_l0_61
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.0.attn.hook_z
    Hook Layer
    0
    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
    0:0.01
    1:0.01
    2:0.02
    3:0.03
    4:0.00
    5:0.01
    6:0.30
    7:0.16
    8:0.01
    9:0.01
    10:0.01
    11:0.00
    12:0.02
    13:0.12
    14:0.20
    15:0.01
    Negative Logits
     autorytatywna
    -0.56
    GEBURTSDATUM
    -0.51
    setVerticalGroup
    -0.51
     betweenstory
    -0.48
     NSCoder
    -0.46
     doInBackground
    -0.43
     ſeine
    -0.43
    expandindo
    -0.43
    ﹧
    -0.42
     ſelbſt
    -0.42
    POSITIVE LOGITS
     laikā
    0.21
     antemano
    0.20
     antaranya
    0.20
     keselamatan
    0.20
     vôtre
    0.20
     egens
    0.20
     nocturno
    0.19
     Handlungen
    0.19
     keempat
    0.19
     dalamnya
    0.18
    Activations Density 0.303%

    No Known Activations