© 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. 5-ATT-KK
    11. 20621
    Prev
    Next
    INDEX
    Explanations

    instances of the word "fact" or phrases that indicate something is being confirmed as true

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 5-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L5_Hcat_z_lr1.20e-03_l11.00e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    Data Type
    float32
    Hook Name
    blocks.5.attn.hook_z
    Hook Layer
    5
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.07
    1:0.04
    2:0.14
    3:0.11
    4:0.06
    5:0.02
    6:0.07
    7:0.11
    8:0.07
    9:0.02
    10:0.04
    11:0.17
    Negative Logits
     congr
    -1.62
    20439
    -1.56
     VIDEOS
    -1.47
    Redditor
    -1.44
    oiler
    -1.39
     synopsis
    -1.39
    版
    -1.37
     commenters
    -1.34
    sonian
    -1.34
     commenter
    -1.34
    POSITIVE LOGITS
     recover
    1.66
     solvent
    1.58
    active
    1.58
    auld
    1.57
    etric
    1.55
     mater
    1.53
    iliated
    1.49
    igmatic
    1.49
    rehend
    1.48
    neutral
    1.48
    Activations Density 0.002%

    No Known Activations