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. 5662
    Prev
    Next
    INDEX
    Explanations

    attends to pronouns referring to groups or individuals from the actions or tokens associated with those pronouns

    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
    0:0.19
    1:0.20
    2:0.33
    3:0.08
    4:0.04
    5:0.01
    6:0.03
    7:0.08
    Negative Logits
    Tikang
    -0.30
     stethoscope
    -0.29
    annica
    -0.29
     patience
    -0.28
     ویکی‌پدیا
    -0.27
    patience
    -0.27
     bParam
    -0.26
    letzt
    -0.26
     defStyleAttr
    -0.26
    cuits
    -0.26
    POSITIVE LOGITS
     themselves
    0.50
    themselves
    0.40
     Audiodateien
    0.34
     themſelves
    0.34
    กัน
    0.32
    AndEndTag
    0.32
     Signalez
    0.31
     själva
    0.31
     توانند
    0.30
     Winaray
    0.29
    Activations Density 0.777%

    No Known Activations