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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Under Peer Review · Attention SAE Research Paper
    3. GPT2-Small
    4. Attention Out
    5. 2-ATT-KK
    6. 22650
    Prev
    Next
    INDEX
    Explanations

    mentions of punk music and related events or content

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

    elements related to punk music culture and its associated genres

    oai_token-act-pair · gpt-4o-miniTriggered 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
    0:0.07
    1:0.06
    2:0.17
    3:0.06
    4:0.08
    5:0.07
    6:0.03
    7:0.04
    8:0.05
    9:0.21
    10:0.07
    11:0.03
    Negative Logits
    ebus
    -1.47
    channelAvailability
    -1.47
    Internal
    -1.43
    andowski
    -1.38
    �士
    -1.38
     deductible
    -1.32
     latitude
    -1.29
     improper
    -1.25
    Downloadha
    -1.24
     probable
    -1.20
    POSITIVE LOGITS
     rag
    1.41
     thriller
    1.39
     slang
    1.39
     classics
    1.38
     punk
    1.38
    punk
    1.36
     comedy
    1.33
    lance
    1.32
     horror
    1.32
    cult
    1.30
    Activations Density 0.015%

    No Known Activations