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. 9-ATT-KK
    6. 18766
    Prev
    Next
    INDEX
    Explanations

    keywords related to conflicts or struggles

    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
    0:0.06
    1:0.05
    2:0.05
    3:0.07
    4:0.24
    5:0.09
    6:0.06
    7:0.13
    8:0.06
    9:0.03
    10:0.05
    11:0.05
    Negative Logits
    rote
    -2.71
    kef
    -2.45
     NCT
    -2.35
    eln
    -2.29
    ucc
    -2.23
    cients
    -2.14
    ataka
    -2.08
    ichick
    -2.05
     PubMed
    -2.03
    uliffe
    -2.00
    POSITIVE LOGITS
     Written
    3.86
     discovers
    2.83
     realizes
    2.76
     reluctantly
    2.65
     convin
    2.59
     confronts
    2.56
     begins
    2.46
     decides
    2.45
     protagonist
    2.44
     ens
    2.41
    Activations Density 0.433%

    No Known Activations