© 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. 4681
    Prev
    Next
    INDEX
    Explanations

    references to important diplomatic visits and agreements

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    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
    0:0.13
    1:0.02
    2:0.26
    3:0.12
    4:0.05
    5:0.12
    6:0.03
    7:0.07
    8:0.06
    9:0.03
    10:0.05
    11:0.03
    Negative Logits
     Crim
    -2.26
     gore
    -2.23
     delinqu
    -2.22
     derivatives
    -2.15
     steril
    -2.10
     clones
    -2.08
     Slayer
    -2.05
    surv
    -2.05
     disenfranch
    -2.04
     amput
    -1.99
    POSITIVE LOGITS
     visit
    3.75
     visits
    3.64
     envoy
    3.25
    meet
    3.22
     bilateral
    3.11
     greet
    3.07
     meetings
    3.04
     visiting
    3.03
     greeting
    3.03
     diplomatic
    3.03
    Activations Density 0.897%

    No Known Activations