Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Under Peer Review · Attention SAE Research Paper
    3. GPT2-Small
    4. Attention Out
    5. 5-ATT-KK
    6. 2693
    Prev
    Next
    INDEX
    Explanations

    the word "but" as a conjunction indicating contrast or an exception

    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
    0:0.02
    1:0.02
    2:0.15
    3:0.05
    4:0.08
    5:0.02
    6:0.11
    7:0.04
    8:0.06
    9:0.05
    10:0.06
    11:0.28
    Negative Logits
     churn
    -1.75
     twists
    -1.69
     timelines
    -1.69
     whirlwind
    -1.65
     sched
    -1.60
     precon
    -1.59
     spins
    -1.59
     horrors
    -1.59
     gestation
    -1.55
     untold
    -1.55
    POSITIVE LOGITS
    etooth
    1.93
    authorized
    1.92
    avorite
    1.86
    allion
    1.82
    abilia
    1.82
    luster
    1.74
    illegal
    1.73
    angible
    1.73
    adra
    1.68
    inguishable
    1.67
    Activations Density 0.001%

    No Known Activations