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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Attention SAE Research Paper
    5. GPT2-Small
    7. Attention Out
    9. 5-ATT-KK
    11. 43170
    Prev
    Next
    INDEX
    Explanations

    repeated patterns and consistent formatting in text

    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
    2:0.05
    3:0.08
    4:0.08
    5:0.03
    6:0.03
    7:0.34
    8:0.03
    9:0.03
    10:0.11
    11:0.13
    Negative Logits
    aceae
    -1.84
     laure
    -1.51
    sonian
    -1.43
    icist
    -1.40
     神
    -1.39
    EMS
    -1.38
    ERA
    -1.36
     Psychiat
    -1.33
    igion
    -1.32
    entric
    -1.31
    POSITIVE LOGITS
     undercover
    1.42
     websites
    1.30
     spoof
    1.28
     situations
    1.25
     unauthorized
    1.23
     embed
    1.21
     encount
    1.21
     kios
    1.20
     encounter
    1.20
     logged
    1.19
    Activations Density 0.003%

    No Known Activations