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. 1199
    Prev
    Next
    INDEX
    Explanations

    attends to nested mathematical expressions on earlier tokens from simpler token structures on later tokens

    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
    2:0.04
    3:0.06
    4:0.00
    5:0.00
    6:0.10
    7:0.07
    8:0.01
    9:0.01
    10:0.01
    11:0.00
    12:0.02
    13:0.04
    14:0.52
    15:0.02
    Negative Logits
    <eos>
    -0.25
    ‚¬
    -0.25
     sebou
    -0.24
    \
    -0.24
    IndentedString
    -0.23
    ">'+
    -0.23
    \
    
    -0.23
     betreft
    -0.23
    }\
    -0.22
    }
    -0.22
    POSITIVE LOGITS
    şört
    0.34
    出版年
    0.32
    ſchen
    0.31
     chande
    0.31
     ſever
    0.31
     Keluarga
    0.31
     يتيمه
    0.31
     للاسماء
    0.31
    ágenes
    0.30
     ſte
    0.30
    Activations Density 0.347%

    No Known Activations