© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Attention SAE Research Paper
    5. GPT2-Small
    7. Attention Out
    9. 10-ATT-KK
    11. 15582
    Prev
    Next
    INDEX
    Explanations

    phrases containing conjunctions, particularly "and" and "or."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L10_Hcat_z_lr1.20e-03_l11.30e+00_ds24576_bs4096_dc1.00e-05_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.10.attn.hook_z
    Hook Layer
    10
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.05
    1:0.03
    2:0.05
    3:0.05
    4:0.22
    5:0.04
    6:0.07
    7:0.22
    8:0.03
    9:0.04
    10:0.03
    11:0.12
    Negative Logits
    gamer
    -3.11
    idepress
    -2.88
    Game
    -2.85
    gam
    -2.66
    GAME
    -2.62
    amaz
    -2.60
    game
    -2.54
    Gamer
    -2.53
     Game
    -2.51
     GAM
    -2.50
    POSITIVE LOGITS
    uba
    2.35
     Hoover
    2.30
     Mayer
    2.27
     Leban
    2.23
     SAR
    2.22
     Lauder
    2.20
    RM
    2.15
     LAW
    2.14
    entials
    2.11
     Pett
    2.11
    Activations Density 0.009%

    No Known Activations