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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 3072
    5. 8-RES_FS3072-JB
    6. 1306
    Prev
    Next
    INDEX
    Explanations

    mentions of legal officials, specifically attorneys general

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs3072-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    3,072
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    571
    +0.14
    0.6%
    679
    +0.13
    0.6%
    730
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    224
    +0.14
    0.06
    571
    +0.13
    0.05
    730
    +0.10
    0.05
    Negative Logits
     Reincarn
    -0.78
    obyl
    -0.76
     Marathon
    -0.65
     celebrations
    -0.65
    omsday
    -0.63
     Delicious
    -0.63
    wn
    -0.62
     Dating
    -0.61
     festive
    -0.61
     Paleo
    -0.60
    POSITIVE LOGITS
     Gloria
    0.97
     solicitor
    0.95
     attorney
    0.93
    orneys
    0.87
     attorneys
    0.83
    maid
    0.81
    yers
    0.81
     lawyer
    0.81
     specializing
    0.79
    suit
    0.79
    Activations Density 0.590%

    No Known Activations