© 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. 2-ATT-KK
    11. 16125
    Prev
    Next
    INDEX
    Explanations

    phrases indicating contrasting or negative outcomes

    oai_token-act-pair · gpt-3.5-turbo
    Triggered 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
    Head Attr Weights
    0:0.12
    1:0.03
    2:0.04
    3:0.25
    4:0.03
    5:0.05
    6:0.05
    7:0.02
    8:0.06
    9:0.16
    10:0.09
    11:0.05
    Negative Logits
    ospons
    -1.45
     toured
    -1.21
     Admission
    -1.18
    LAN
    -1.16
    Board
    -1.13
    intendent
    -1.12
    ullivan
    -1.11
    ahime
    -1.08
    ortunately
    -1.05
    everal
    -1.04
    POSITIVE LOGITS
     nor
    1.68
     whatsoever
    1.40
    nor
    1.36
     Pik
    1.23
    ifice
    1.21
     spark
    1.19
     slightest
    1.18
     necessarily
    1.18
    isable
    1.14
     anymore
    1.10
    Activations Density 0.076%

    No Known Activations