© 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. 5-ATT-KK
    11. 6205
    Prev
    Next
    INDEX
    Explanations

    instances of misunderstanding or misinterpretation

    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
    0:0.01
    1:0.00
    2:0.07
    3:0.05
    4:0.15
    5:0.03
    6:0.04
    7:0.40
    8:0.02
    9:0.03
    10:0.08
    11:0.07
    Negative Logits
    ectomy
    -1.61
    iann
    -1.46
     supervised
    -1.39
    umar
    -1.39
    idal
    -1.38
    cum
    -1.37
    cone
    -1.37
    inth
    -1.37
    urable
    -1.35
    tar
    -1.33
    POSITIVE LOGITS
     misunderstand
    1.76
     misunderstood
    1.54
     misunderstanding
    1.42
     mistake
    1.40
     mistakes
    1.40
     miscon
    1.38
    assetsadobe
    1.33
     unprepared
    1.33
     glances
    1.32
     terminology
    1.32
    Activations Density 0.019%

    No Known Activations