© 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. 3-ATT-KK
    11. 13394
    Prev
    Next
    INDEX
    Explanations

    references to behind-the-scenes activities or events

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

    mentions of behind-the-scenes content

    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
    1:0.03
    2:0.09
    3:0.20
    4:0.01
    5:0.03
    6:0.11
    7:0.10
    8:0.06
    9:0.12
    10:0.08
    11:0.09
    Negative Logits
    cit
    -1.43
    obar
    -1.39
    inia
    -1.38
    respective
    -1.34
    urdue
    -1.31
     sqor
    -1.29
    ipped
    -1.26
    orically
    -1.26
    inge
    -1.25
    ongyang
    -1.23
    POSITIVE LOGITS
    Redditor
    1.25
     renaissance
    1.16
     guide
    1.15
    Appearances
    1.14
     blueprint
    1.13
     documentary
    1.10
     revival
    1.09
    Alias
    1.08
    タ
    1.08
     template
    1.07
    Activations Density 0.002%

    No Known Activations