© 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. 10354
    Prev
    Next
    INDEX
    Explanations

    phrases related to the act of negating or dismissing concepts and arguments

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 5-att-kk
    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
    0:0.02
    1:0.02
    2:0.04
    3:0.08
    4:0.19
    5:0.04
    6:0.02
    7:0.33
    8:0.04
    9:0.05
    10:0.06
    11:0.06
    Negative Logits
    HEAD
    -1.77
    PHOTOS
    -1.74
     Thumbnails
    -1.73
    代
    -1.67
    人
    -1.64
    CEO
    -1.60
    ocamp
    -1.58
    Screenshot
    -1.57
    umbn
    -1.57
    Leaks
    -1.56
    POSITIVE LOGITS
     inhibition
    1.75
     inhib
    1.71
     possibility
    1.71
     outright
    1.71
     altogether
    1.70
     contention
    1.69
     prejudice
    1.68
     tampering
    1.66
     causation
    1.64
     prevailing
    1.62
    Activations Density 0.001%

    No Known Activations