© 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. 2-ATT-KK
    11. 14899
    Prev
    Next
    INDEX
    Explanations

    phrases related to official statements

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L2_Hcat_z_lr1.20e-03_l11.00e+00_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v4.pt
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    float32
    Hook Name
    blocks.2.attn.hook_z
    Hook Layer
    2
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.07
    1:0.04
    2:0.12
    3:0.13
    4:0.05
    5:0.07
    6:0.07
    7:0.03
    8:0.16
    9:0.10
    10:0.09
    11:0.03
    Negative Logits
     brakes
    -1.52
     Blacks
    -1.51
    rogens
    -1.49
     markers
    -1.47
     invaders
    -1.47
     sinners
    -1.47
     treasures
    -1.46
     obstacles
    -1.44
     staples
    -1.44
     favourites
    -1.43
    POSITIVE LOGITS
     context
    1.73
    statement
    1.64
     briefing
    1.59
     manner
    1.55
    ibus
    1.52
    ICLE
    1.51
     courtyard
    1.41
     courtroom
    1.41
     formulation
    1.39
    javascript
    1.39
    Activations Density 0.078%

    No Known Activations