© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Attention SAE Research Paper
    5. GPT2-Small
    7. Attention Out
    9. 10-ATT-KK
    11. 11950
    Prev
    Next
    INDEX
    Explanations

    emotional responses and descriptors related to negative or judgmental situations

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L10_Hcat_z_lr1.20e-03_l11.30e+00_ds24576_bs4096_dc1.00e-05_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.10.attn.hook_z
    Hook Layer
    10
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Head Attr Weights
    0:0.02
    1:0.02
    2:0.02
    3:0.03
    4:0.04
    5:0.08
    6:0.02
    7:0.03
    8:0.01
    9:0.60
    10:0.02
    11:0.06
    Negative Logits
    TING
    -4.25
    pleting
    -4.13
    angering
    -4.06
    forming
    -3.96
    thening
    -3.95
    joining
    -3.93
     forming
    -3.92
    roying
    -3.91
    ving
    -3.86
     releasing
    -3.86
    POSITIVE LOGITS
    scribe
    2.60
    rued
    2.37
    rahim
    2.22
    abus
    2.13
    SPONSORED
    2.12
    PLIED
    2.09
     Freak
    2.00
    verages
    1.98
     dictated
    1.97
    urned
    1.97
    Activations Density 0.503%

    No Known Activations