© 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. 9-ATT-KK
    11. 3890
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    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
    0:0.09
    1:0.04
    2:0.03
    3:0.09
    4:0.09
    5:0.08
    6:0.05
    7:0.19
    8:0.06
    9:0.03
    10:0.07
    11:0.12
    Negative Logits
     ­
    -2.96
    )—
    -2.92
    "—
    -2.78
     […]
    -2.48
    —"
    -2.36
    —
    -2.25
     tranqu
    -2.24
     bluff
    -2.23
     ∼
    -2.19
     territ
    -2.17
    POSITIVE LOGITS
    enegger
    3.28
    mens
    2.90
    idth
    2.88
    ................
    2.67
    ........
    2.59
    Posted
    2.57
    igrants
    2.50
    acci
    2.50
    lein
    2.45
    obin
    2.38
    Activations Density 0.000%

    No Known Activations

    This feature has no known activations.