© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 5-ATT-KK
    11. 44786
    Prev
    Next
    INDEX
    Explanations

    references to positions of power or authority within government

    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
    1:0.01
    2:0.16
    3:0.04
    4:0.17
    5:0.06
    6:0.04
    7:0.13
    8:0.05
    9:0.04
    10:0.11
    11:0.10
    Negative Logits
     blogs
    -1.50
     subreddits
    -1.49
     sock
    -1.47
     decks
    -1.43
    class
    -1.41
     Markets
    -1.40
     journals
    -1.38
     sym
    -1.38
     file
    -1.37
     precincts
    -1.36
    POSITIVE LOGITS
    irlf
    1.72
    hetamine
    1.53
    llan
    1.52
    oğan
    1.47
    WAY
    1.44
    roximately
    1.43
    ________________________________
    1.42
    Attempt
    1.42
    omore
    1.41
    urther
    1.40
    Activations Density 0.001%

    No Known Activations