© 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. 5-ATT-KK
    11. 25342
    Prev
    Next
    INDEX
    Explanations

    adjectives and adjectives ending in 'ed' that increase interest or curiosity

    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
    2:0.07
    3:0.05
    4:0.22
    5:0.01
    6:0.05
    7:0.37
    8:0.02
    9:0.02
    10:0.03
    11:0.06
    Negative Logits
    liv
    -1.76
    ogether
    -1.66
     uninsured
    -1.63
     numb
    -1.56
    ectomy
    -1.53
    rition
    -1.49
    ops
    -1.48
    clerosis
    -1.48
    ���
    -1.46
     commute
    -1.45
    POSITIVE LOGITS
     popularity
    1.91
     ingenuity
    1.66
     greatness
    1.65
     curiosity
    1.62
     fame
    1.59
     rhetorical
    1.58
     Enlightenment
    1.57
     Aph
    1.50
     publicity
    1.50
     lofty
    1.48
    Activations Density 0.000%

    No Known Activations