© 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. 3-ATT-KK
    11. 13309
    Prev
    Next
    INDEX
    Explanations

    specific instances referring to what something said or information provided

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot

    the conjunction "that" in various contexts, often related to reporting or stating situations or beliefs

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L3_Hcat_z_lr1.20e-03_l19.00e-01_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.3.attn.hook_z
    Hook Layer
    3
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.04
    1:0.05
    2:0.09
    3:0.04
    4:0.03
    5:0.08
    6:0.08
    7:0.07
    8:0.19
    9:0.11
    10:0.11
    11:0.06
    Negative Logits
    equ
    -1.21
    river
    -1.12
    eta
    -1.09
    hattan
    -1.09
    エ
    -1.04
    weights
    -1.01
    anger
    -1.01
    tails
    -1.00
    Atlantic
    -0.99
    2200
    -0.98
    POSITIVE LOGITS
     "...
    1.23
    perty
    1.23
     suscept
    1.15
     subsequ
    1.12
    owship
    1.10
     discrep
    1.08
    izoph
    1.05
    FTWARE
    1.05
     photograp
    1.04
     plagiar
    1.02
    Activations Density 0.098%

    No Known Activations