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

    Neuronpedia

    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. 2-ATT-KK
    11. 22963
    Prev
    Next
    INDEX
    Explanations

    educational institutions and potential events or announcements at those institutions

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

    mentions of hope and associated sentiments

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L2_Hcat_z_lr1.20e-03_l11.00e+00_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v4.pt
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    float32
    Hook Name
    blocks.2.attn.hook_z
    Hook Layer
    2
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.06
    1:0.03
    2:0.14
    3:0.12
    4:0.15
    5:0.06
    6:0.03
    7:0.02
    8:0.10
    9:0.15
    10:0.05
    11:0.03
    Negative Logits
    translation
    -1.22
     accounted
    -1.19
    ital
    -1.17
    account
    -1.14
    jab
    -1.13
    う
    -1.12
    ictionary
    -1.11
    polit
    -1.11
    Statistics
    -1.11
    annot
    -1.09
    POSITIVE LOGITS
    aukee
    1.30
    amaz
    1.29
     Pavilion
    1.16
    amen
    1.16
    oria
    1.15
     Parenthood
    1.13
     Peak
    1.10
     Reloaded
    1.09
     Apollo
    1.09
    sworth
    1.09
    Activations Density 0.003%

    No Known Activations