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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Attention SAE Research Paper
    5. GPT2-Small
    7. Attention Out
    9. 10-ATT-KK
    11. 22403
    Prev
    Next
    INDEX
    Explanations

    phrases that indicate violations of trust and policies in a professional context

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L10_Hcat_z_lr1.20e-03_l11.30e+00_ds24576_bs4096_dc1.00e-05_rsanthropic_rie25000_nr4_v9.pt
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    float32
    Hook Name
    blocks.10.attn.hook_z
    Hook Layer
    10
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.07
    1:0.07
    2:0.16
    3:0.12
    4:0.03
    5:0.01
    6:0.08
    7:0.09
    8:0.03
    9:0.03
    10:0.22
    11:0.04
    Negative Logits
     Norn
    -3.54
     Ily
    -3.25
     Fors
    -3.20
     Vor
    -3.11
    ¯
    -2.95
    Hen
    -2.86
     Hen
    -2.82
     Zup
    -2.78
     guiActiveUn
    -2.77
    ェ
    -2.76
    POSITIVE LOGITS
     AM
    10.00
    AM
    6.90
     PM
    6.62
    PM
    5.34
    AMS
    5.09
     Am
    4.92
     FM
    4.70
    AMI
    4.65
    AMP
    4.49
     HAM
    4.46
    Activations Density 0.025%

    No Known Activations