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

    Neuronpedia

    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. 35262
    Prev
    Next
    INDEX
    Explanations

    elevated emotional expressions, particularly frustration or intensity in interactions

    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
    0:0.21
    1:0.04
    2:0.04
    3:0.06
    4:0.05
    5:0.20
    6:0.03
    7:0.03
    8:0.10
    9:0.05
    10:0.05
    11:0.09
    Negative Logits
    chrom
    -2.07
    stre
    -2.07
    cell
    -2.02
    mes
    -2.01
    )</
    -2.01
    card
    -2.01
    string
    -1.99
    elo
    -1.99
    sole
    -1.98
    usc
    -1.96
    POSITIVE LOGITS
    uyomi
    1.79
     Wonderland
    1.72
     Wealth
    1.67
     Nightmares
    1.67
     Witches
    1.59
     Sheep
    1.58
     Q
    1.58
     Dud
    1.54
     Dice
    1.54
     Ill
    1.54
    Activations Density 0.004%

    No Known Activations