© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Attention SAE Research Paper
    5. GPT2-Small
    7. Attention Out
    9. 9-ATT-KK
    11. 789
    Prev
    Next
    INDEX
    Explanations

    references to notable achievements or significant accomplishments

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 9-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L9_Hcat_z_lr1.20e-03_l11.20e+00_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    float32
    Hook Name
    blocks.9.attn.hook_z
    Hook Layer
    9
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.07
    1:0.07
    2:0.09
    3:0.08
    4:0.08
    5:0.06
    6:0.07
    7:0.09
    8:0.08
    9:0.08
    10:0.08
    11:0.09
    Negative Logits
    teenth
    -3.05
     Nanto
    -2.84
    ento
    -2.79
    Favorite
    -2.62
    itsu
    -2.54
    itsch
    -2.52
    iotics
    -2.51
    rade
    -2.50
    iquid
    -2.47
    hearted
    -2.46
    POSITIVE LOGITS
     eleph
    2.89
     Divide
    2.81
     Palest
    2.71
     cascade
    2.67
     ‎
    2.67
     Bore
    2.56
     Cascade
    2.54
     glac
    2.51
     Kern
    2.50
     campaigner
    2.48
    Activations Density 0.000%

    No Known Activations