© 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. 38074
    Prev
    Next
    INDEX
    Explanations

    key pronouns and verbs in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered 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
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.02
    1:0.03
    2:0.04
    3:0.19
    4:0.25
    5:0.05
    6:0.09
    7:0.05
    8:0.04
    9:0.07
    10:0.05
    11:0.06
    Negative Logits
    emale
    -1.66
    adelphia
    -1.55
     Palest
    -1.44
    %);
    -1.43
     Ascension
    -1.42
    undrum
    -1.39
     cradle
    -1.39
     Bulgar
    -1.38
     Pwr
    -1.38
    ylon
    -1.37
    POSITIVE LOGITS
     accordingly
    1.49
     Quote
    1.40
     wouldn
    1.36
     bets
    1.34
     shouldn
    1.34
    llular
    1.32
     guess
    1.31
     devs
    1.30
     refresh
    1.30
    シャ
    1.29
    Activations Density 0.100%

    No Known Activations