© 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. 3-ATT-KK
    11. 12989
    Prev
    Next
    INDEX
    Explanations

    internal links to various content throughout the text

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot

    phrases relating to important updates or announcements

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L3_Hcat_z_lr1.20e-03_l19.00e-01_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.3.attn.hook_z
    Hook Layer
    3
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.01
    1:0.10
    2:0.15
    3:0.04
    4:0.03
    5:0.03
    6:0.22
    7:0.12
    8:0.07
    9:0.06
    10:0.06
    11:0.06
    Negative Logits
     heck
    -0.98
    harm
    -0.97
     verify
    -0.95
     SIGN
    -0.93
     bills
    -0.91
    bered
    -0.89
    ."
    -0.89
     pretend
    -0.88
    reaching
    -0.87
     downstream
    -0.85
    POSITIVE LOGITS
    ================================================================
    1.20
    ippery
    1.07
    erker
    1.00
    sonian
    1.00
    iaz
    1.00
    riot
    1.00
    Products
    0.99
    ========
    0.98
    enza
    0.96
    oggles
    0.96
    Activations Density 0.245%

    No Known Activations