© 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. 9-ATT-KK
    11. 1628
    Prev
    Next
    INDEX
    Explanations

    phrases indicating presence or absence of individuals

    oai_token-act-pair · gpt-4o-miniTriggered 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
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.13
    1:0.03
    2:0.16
    3:0.13
    4:0.06
    5:0.12
    6:0.06
    7:0.07
    8:0.08
    9:0.04
    10:0.04
    11:0.04
    Negative Logits
     responding
    -2.62
     suspected
    -2.43
     reports
    -2.36
     investigators
    -2.31
     surveillance
    -2.31
     FBI
    -2.30
     reporting
    -2.26
     suspects
    -2.26
     retaliation
    -2.25
     targeted
    -2.25
    POSITIVE LOGITS
    agame
    2.84
     サーティワン
    2.60
    olkien
    2.39
    ciation
    2.37
    natureconservancy
    2.36
    ructose
    2.30
    MpServer
    2.30
     Flavoring
    2.29
    byss
    2.25
    gebra
    2.23
    Activations Density 0.002%

    No Known Activations