© 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. 12275
    Prev
    Next
    INDEX
    Explanations

    instances where positive qualities or achievements are being described

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

    instances of contrast and comparison in narratives

    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
    0:0.02
    1:0.04
    2:0.07
    3:0.03
    4:0.03
    5:0.08
    6:0.09
    7:0.28
    8:0.07
    9:0.04
    10:0.08
    11:0.12
    Negative Logits
    FTWARE
    -1.59
    ゴン
    -1.58
    appa
    -1.48
     sidx
    -1.46
     sqor
    -1.39
     Activate
    -1.37
    Reloaded
    -1.35
    Newsletter
    -1.30
    DATA
    -1.30
     arrang
    -1.29
    POSITIVE LOGITS
    pires
    1.37
     awaits
    1.34
     arose
    1.31
     looms
    1.30
     ripe
    1.30
     dominated
    1.24
    dominated
    1.22
     inhabited
    1.19
     arises
    1.18
    imming
    1.18
    Activations Density 0.581%

    No Known Activations