© 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. 2-ATT-KK
    11. 14314
    Prev
    Next
    INDEX
    Explanations

    names of plants

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L2_Hcat_z_lr1.20e-03_l11.00e+00_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v4.pt
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    float32
    Hook Name
    blocks.2.attn.hook_z
    Hook Layer
    2
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.06
    1:0.04
    2:0.20
    3:0.05
    4:0.22
    5:0.06
    6:0.04
    7:0.03
    8:0.04
    9:0.09
    10:0.08
    11:0.03
    Negative Logits
     unborn
    -1.49
    rint
    -1.30
    thood
    -1.29
     diagn
    -1.26
     induct
    -1.25
    ikuman
    -1.23
     cures
    -1.18
    mates
    -1.18
     UID
    -1.18
     jurors
    -1.18
    POSITIVE LOGITS
    ascus
    1.54
    ioxide
    1.41
     Caucasus
    1.41
     Tig
    1.33
    orus
    1.27
    idated
    1.23
    rika
    1.22
     PsyNet
    1.19
     Valley
    1.18
     Shal
    1.17
    Activations Density 0.002%

    No Known Activations