© 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. 33502
    Prev
    Next
    INDEX
    Explanations

    derogatory and disparaging language toward specific ethnic or religious groups

    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
    0:0.01
    1:0.00
    2:0.14
    3:0.08
    4:0.11
    5:0.02
    6:0.03
    7:0.13
    8:0.04
    9:0.04
    10:0.17
    11:0.14
    Negative Logits
     CLIENT
    -1.25
    )].
    -1.22
     largeDownload
    -1.16
    orem
    -1.13
    psc
    -1.07
     miracle
    -1.07
     Heal
    -1.03
    ettle
    -1.02
    imester
    -1.01
     Agric
    -1.00
    POSITIVE LOGITS
     targets
    1.14
     targeting
    1.12
    rex
    1.10
     imaginable
    1.09
     raping
    1.09
     Adolf
    1.07
    tones
    1.05
     outlawed
    1.04
     pedoph
    1.02
     surfaced
    1.02
    Activations Density 0.106%

    No Known Activations