© 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
    2. Under Peer Review · Attention SAE Research Paper
    3. GPT2-Small
    4. Attention Out
    5. 10-ATT-KK
    6. 7236
    Prev
    Next
    INDEX
    Explanations

    phrases that critique societal attitudes and practices regarding morality and legislation

    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
    0:0.13
    1:0.01
    2:0.13
    3:0.03
    4:0.04
    5:0.37
    6:0.01
    7:0.06
    8:0.02
    9:0.07
    10:0.04
    11:0.02
    Negative Logits
    rather
    -4.25
     rather
    -3.99
     instead
    -3.93
    instead
    -3.49
    stice
    -2.98
    appa
    -2.92
    jected
    -2.89
    artz
    -2.87
     later
    -2.78
     earlier
    -2.73
    POSITIVE LOGITS
     anymore
    5.95
     nor
    4.24
    yet
    3.94
    nor
    3.47
     unless
    3.43
     either
    3.34
    unless
    3.31
     necessarily
    3.16
    ¨
    3.13
    �
    3.12
    Activations Density 0.456%

    No Known Activations