© 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. 5-ATT-KK
    11. 38282
    Prev
    Next
    INDEX
    Explanations

    phrases that imply deception or dishonesty

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L5_Hcat_z_lr1.20e-03_l11.00e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    Data Type
    float32
    Hook Name
    blocks.5.attn.hook_z
    Hook Layer
    5
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Head Attr Weights
    0:0.26
    1:0.09
    2:0.02
    3:0.03
    4:0.02
    5:0.27
    6:0.10
    7:0.01
    8:0.08
    9:0.04
    10:0.02
    11:0.01
    Negative Logits
     Codec
    -2.18
     Rolls
    -1.70
     Lich
    -1.65
     precinct
    -1.63
     {:
    -1.62
     Layer
    -1.60
     Singer
    -1.60
     SOFTWARE
    -1.59
     Berkshire
    -1.57
     Scroll
    -1.56
    POSITIVE LOGITS
    agree
    2.63
    mad
    2.58
    ped
    2.43
    d
    2.37
    df
    2.34
    reed
    2.33
    few
    2.33
    ende
    2.31
    reb
    2.27
    bably
    2.27
    Activations Density 0.020%

    No Known Activations