© 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. 3-ATT-KK
    11. 11827
    Prev
    Next
    INDEX
    Explanations

    phrases related to the use of certain substances or items, specifically related to legal, medical, or environmental contexts

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot

    phrases emphasizing the concept of "of" relating to various subjects and objects

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L3_Hcat_z_lr1.20e-03_l19.00e-01_ds24576_bs4096_dc1.00e-06_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.3.attn.hook_z
    Hook Layer
    3
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.01
    1:0.02
    2:0.06
    3:0.28
    4:0.01
    5:0.02
    6:0.08
    7:0.18
    8:0.06
    9:0.06
    10:0.09
    11:0.08
    Negative Logits
    NR
    -1.27
    atform
    -1.27
    ukemia
    -1.19
    stellar
    -1.16
    ences
    -1.15
    earchers
    -1.14
    dq
    -1.08
    clerosis
    -1.08
    uzzle
    -1.07
    ector
    -1.07
    POSITIVE LOGITS
    ダ
    1.29
     Upton
    1.23
     conditioning
    1.23
    ザ
    1.22
     ART
    1.21
     IPM
    1.18
     awa
    1.17
    ause
    1.13
    aukee
    1.11
     undue
    1.10
    Activations Density 0.011%

    No Known Activations