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. 9402
    Prev
    Next
    INDEX
    Explanations

    instances of refusal or resistance

    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
    0:0.08
    1:0.08
    2:0.06
    3:0.14
    4:0.06
    5:0.09
    6:0.08
    7:0.06
    8:0.08
    9:0.12
    10:0.04
    11:0.04
    Negative Logits
    earance
    -1.99
    him
    -1.98
     condol
    -1.97
     disple
    -1.97
     disagree
    -1.96
     disagrees
    -1.89
    Neither
    -1.89
     uncond
    -1.87
    agree
    -1.86
     refuse
    -1.84
    POSITIVE LOGITS
     macros
    2.07
     multiplier
    2.05
     generators
    1.98
    soDeliveryDate
    1.94
    nets
    1.87
     analytics
    1.85
     initiation
    1.79
     stakes
    1.78
     mete
    1.78
     hackers
    1.78
    Activations Density 0.001%

    No Known Activations