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. 7-ATT-KK
    6. 9251
    Prev
    Next
    INDEX
    Explanations

    incidents involving physical violence or criminal activity

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 7-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L7_Hcat_z_lr1.20e-03_l11.10e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    Data Type
    float32
    Hook Name
    blocks.7.attn.hook_z
    Hook Layer
    7
    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
    1:0.20
    2:0.02
    3:0.10
    4:0.07
    5:0.07
    6:0.09
    7:0.04
    8:0.06
    9:0.10
    10:0.07
    11:0.05
    Negative Logits
    cision
    -2.47
     Trials
    -2.00
    Dub
    -1.98
     miracles
    -1.92
    isSpecialOrderable
    -1.88
     Ashes
    -1.86
    optim
    -1.85
     immortality
    -1.85
     2022
    -1.85
     Exile
    -1.82
    POSITIVE LOGITS
     tresp
    2.30
     stationary
    2.23
     disorderly
    2.14
     disturbance
    2.03
     parked
    1.98
    EMS
    1.96
     escalated
    1.95
     occupants
    1.94
     Pastebin
    1.92
     intoxicated
    1.91
    Activations Density 0.106%

    No Known Activations