© 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. 3-ATT-KK
    11. 12453
    Prev
    Next
    INDEX
    Explanations

    sentences containing terms related to social events, news, and reports

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

    phrases or terms related to institutional or bureaucratic critiques

    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
    0:0.03
    1:0.17
    2:0.11
    3:0.04
    4:0.04
    5:0.08
    6:0.14
    7:0.08
    8:0.07
    9:0.08
    10:0.06
    11:0.06
    Negative Logits
    interstitial
    -2.03
    gerald
    -1.67
     outwe
    -1.46
    luster
    -1.36
    interrupted
    -1.34
    etheless
    -1.32
     conclud
    -1.32
     blat
    -1.31
    renheit
    -1.31
    avorite
    -1.28
    POSITIVE LOGITS
     �
    1.59
     �
    1.46
     �
    1.33
     Qin
    1.33
     �
    1.32
     �
    1.27
     �
    1.27
     �
    1.26
     �
    1.26
     Dia
    1.24
    Activations Density 0.076%

    No Known Activations