© 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. 5840
    Prev
    Next
    INDEX
    Explanations

    affirmations or emphatic statements, particularly those expressing opinions or beliefs

    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
    1:0.05
    2:0.14
    3:0.08
    4:0.01
    5:0.04
    6:0.06
    7:0.13
    8:0.10
    9:0.10
    10:0.07
    11:0.14
    Negative Logits
     allegedly
    -1.08
     dise
    -1.02
     supposedly
    -0.98
    Keefe
    -0.95
     Maver
    -0.93
     Grant
    -0.87
     Ree
    -0.87
     Reb
    -0.85
     Dise
    -0.84
    videos
    -0.82
    POSITIVE LOGITS
    stan
    1.24
     caveat
    1.24
     anecd
    1.07
    Background
    1.06
     clarification
    1.02
    ozy
    1.02
    nostic
    0.99
     caveats
    0.98
     subp
    0.98
     goodbye
    0.96
    Activations Density 0.026%

    No Known Activations