© 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. 12653
    Prev
    Next
    INDEX
    Explanations

    references to the Who band and its association with various elements in the text

    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
    0:0.04
    1:0.13
    2:0.05
    3:0.06
    4:0.07
    5:0.07
    6:0.06
    7:0.08
    8:0.21
    9:0.07
    10:0.05
    11:0.07
    Negative Logits
     mates
    -2.46
     then
    -2.44
     GOODMAN
    -2.40
     basically
    -2.18
    actually
    -2.16
     actually
    -2.16
     Totem
    -2.15
     Horizons
    -2.15
     Asc
    -2.10
     hereby
    -2.09
    POSITIVE LOGITS
    achus
    2.58
    efe
    2.56
    cedented
    2.52
    airo
    2.48
    phabet
    2.45
    ャ
    2.42
    nown
    2.36
     WATCHED
    2.33
    ット
    2.32
    plete
    2.29
    Activations Density 0.001%

    No Known Activations