Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Under Peer Review · Attention SAE Research Paper
    3. GPT2-Small
    4. Attention Out
    5. 2-ATT-KK
    6. 13762
    Prev
    Next
    INDEX
    Explanations

    proper names with a specific pattern: "amb" followed by "aram" or "more," along with a separate activation for the sequence "id p on."

    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
    0:0.06
    1:0.03
    2:0.26
    3:0.04
    4:0.09
    5:0.08
    6:0.03
    7:0.02
    8:0.09
    9:0.16
    10:0.06
    11:0.03
    Negative Logits
     conclud
    -1.43
    Bench
    -1.39
     ACTIONS
    -1.34
    inelli
    -1.26
    idon
    -1.21
    wagen
    -1.20
    π
    -1.18
    Thumbnail
    -1.17
     sidelines
    -1.16
    Reviewed
    -1.16
    POSITIVE LOGITS
    REDACTED
    1.30
    lease
    1.13
    vae
    1.08
    criptions
    1.07
    tnc
    1.06
    emon
    1.03
    stadt
    1.02
     Ae
    1.02
    otech
    1.01
     RFC
    1.01
    Activations Density 0.028%

    No Known Activations