© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Attention SAE Research Paper
    5. GPT2-Small
    7. Attention Out
    9. 10-ATT-KK
    11. 6569
    Prev
    Next
    INDEX
    Explanations

    references to cheating and accountability in various contexts

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L10_Hcat_z_lr1.20e-03_l11.30e+00_ds24576_bs4096_dc1.00e-05_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.10.attn.hook_z
    Hook Layer
    10
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Head Attr Weights
    0:0.16
    1:0.07
    2:0.05
    3:0.07
    4:0.10
    5:0.03
    6:0.12
    7:0.08
    8:0.03
    9:0.05
    10:0.12
    11:0.06
    Negative Logits
     Huss
    -3.43
    phabet
    -3.06
     Moose
    -2.90
    Jar
    -2.90
     Ragnar
    -2.87
     Dickinson
    -2.78
     Rochester
    -2.73
    xtap
    -2.69
     Philippe
    -2.67
    dyl
    -2.67
    POSITIVE LOGITS
     cheating
    7.41
     cheat
    7.09
     cheated
    5.88
     che
    5.50
    cheat
    5.13
     gam
    3.83
     blackmail
    3.69
     doping
    3.68
     adultery
    3.47
    gam
    3.40
    Activations Density 0.011%

    No Known Activations