© 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. 3391
    Prev
    Next
    INDEX
    Explanations

    phrases that involve comparison and contrast

    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
    0:0.05
    1:0.31
    2:0.05
    3:0.04
    4:0.04
    5:0.19
    6:0.04
    7:0.02
    8:0.06
    9:0.04
    10:0.06
    11:0.04
    Negative Logits
    PB
    -2.06
    mot
    -1.81
    MAP
    -1.72
    HK
    -1.70
     PsyNet
    -1.68
    Stephen
    -1.67
     Mushroom
    -1.56
    Works
    -1.56
     Helic
    -1.54
    BIP
    -1.53
    POSITIVE LOGITS
    uries
    1.89
    hran
    1.88
    aceous
    1.72
    rity
    1.65
    ihar
    1.61
    ane
    1.61
    aste
    1.60
     Dane
    1.58
     ash
    1.57
     doub
    1.57
    Activations Density 0.001%

    No Known Activations