© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Attention SAE Research Paper
    5. GPT2-Small
    7. Attention Out
    9. 9-ATT-KK
    11. 21948
    Prev
    Next
    INDEX
    Explanations

    phrases about damage or harm done by individuals or groups over time

    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
    0:0.35
    1:0.01
    2:0.26
    3:0.06
    4:0.01
    5:0.04
    6:0.01
    7:0.03
    8:0.03
    9:0.01
    10:0.11
    11:0.02
    Negative Logits
     Airways
    -2.44
     introductory
    -2.40
     answ
    -2.34
     quiz
    -2.29
     chatting
    -2.28
     ticket
    -2.25
     Pilot
    -2.23
     answering
    -2.23
     Answers
    -2.22
    ptin
    -2.20
    POSITIVE LOGITS
     damages
    4.10
    damage
    4.03
     damage
    3.83
     inflicted
    3.77
     Damage
    3.50
    Dam
    3.40
    dam
    3.39
     DAM
    3.20
     repaired
    3.18
     damaging
    3.16
    Activations Density 0.034%

    No Known Activations