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. 1290
    Prev
    Next
    INDEX
    Explanations

    attends to tokens expressing the concept of "already" from later tokens that indicate a prior occurrence or established state

    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
    0:0.02
    1:0.02
    2:0.02
    3:0.04
    4:0.02
    5:0.02
    6:0.08
    7:0.07
    8:0.01
    9:0.01
    10:0.01
    11:0.07
    12:0.02
    13:0.24
    14:0.26
    15:0.02
    Negative Logits
     antidesliz
    -0.28
     vecino
    -0.27
     almofada
    -0.27
     razonable
    -0.26
     monasterio
    -0.26
     createSlice
    -0.25
     telefónica
    -0.25
     húmedo
    -0.25
     tombé
    -0.25
     Absicht
    -0.25
    POSITIVE LOGITS
    cola
    0.33
     cid
    0.33
     cime
    0.32
     Modifier
    0.31
    sure
    0.31
     Dire
    0.31
     jsPsych
    0.31
     georg
    0.31
    cir
    0.31
    bite
    0.31
    Activations Density 0.078%

    No Known Activations