© 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. 42503
    Prev
    Next
    INDEX
    Explanations

    phrases that include combinations of the words "and" or "true" with negative prefixes

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    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
    1:0.01
    2:0.07
    3:0.07
    4:0.06
    5:0.04
    6:0.33
    7:0.15
    8:0.05
    9:0.05
    10:0.07
    11:0.05
    Negative Logits
    VERTISEMENT
    -1.43
    ********************************
    -1.30
     Parables
    -1.29
    ellar
    -1.26
     Hert
    -1.24
     7000
    -1.20
     Mood
    -1.19
     Digest
    -1.16
     "$:/
    -1.15
     Enhancement
    -1.14
    POSITIVE LOGITS
     prosec
    1.42
    tal
    1.33
    ilings
    1.27
    puff
    1.27
    ciples
    1.27
     indict
    1.26
    encers
    1.25
    iaries
    1.24
     outl
    1.23
    aughs
    1.22
    Activations Density 0.001%

    No Known Activations