© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Attention SAE Research Paper
    5. GPT2-Small
    7. Attention Out
    9. 9-ATT-KK
    11. 6357
    Prev
    Next
    INDEX
    Explanations

    phrases that indicate the importance and relevance of main stories or topics

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 9-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L9_Hcat_z_lr1.20e-03_l11.20e+00_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.9.attn.hook_z
    Hook Layer
    9
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Head Attr Weights
    0:0.24
    1:0.01
    2:0.03
    3:0.33
    4:0.02
    5:0.07
    6:0.02
    7:0.08
    8:0.04
    9:0.01
    10:0.07
    11:0.02
    Negative Logits
     insurers
    -2.44
     bureaucrats
    -2.44
    Reviewer
    -2.30
     therapists
    -2.28
     lobbyists
    -2.27
     Consumers
    -2.26
     landlords
    -2.25
     brokers
    -2.24
     consultants
    -2.23
     ISPs
    -2.21
    POSITIVE LOGITS
    utorial
    2.41
     script
    2.23
    reath
    2.22
    avia
    2.15
    elfth
    2.11
    iHUD
    2.05
    ettings
    2.02
    script
    2.01
    estyle
    1.96
    ascript
    1.93
    Activations Density 0.003%

    No Known Activations