© 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 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. 3677
    Prev
    Next
    INDEX
    Explanations

    references to victims or harm in various contexts

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

    No Comments

    Head Attr Weights
    0:0.02
    1:0.02
    2:0.13
    3:0.09
    4:0.33
    5:0.03
    6:0.06
    7:0.12
    8:0.02
    9:0.03
    10:0.05
    11:0.03
    Negative Logits
    awaru
    -1.76
    wcsstore
    -1.62
    TION
    -1.60
    quickShipAvailable
    -1.46
     proponents
    -1.41
     Cosponsors
    -1.33
     Teach
    -1.30
    rase
    -1.29
     Finder
    -1.29
     correctly
    -1.28
    POSITIVE LOGITS
    龍喚士
    1.74
    bowl
    1.63
     doorstep
    1.63
    sic
    1.63
     pavement
    1.61
     driveway
    1.52
     pandemonium
    1.46
    boards
    1.46
     sidewalk
    1.44
     littered
    1.42
    Activations Density 0.008%

    No Known Activations