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

    Neuronpedia

    APISteerSAE EvalsBlog/PodcastNEWSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlog/PodcastGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 3072
    5. 8-RES_FS3072-JB
    6. 1184
    Prev
    Next
    INDEX
    Explanations

    the names of various famous people or titles of events

    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
    362
    +0.11
    0.5%
    699
    +0.11
    0.5%
    506
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    752
    +0.11
    0.06
    506
    +0.11
    0.06
    699
    +0.10
    0.04
    Negative Logits
    otle
    -0.81
    anguage
    -0.80
    ramid
    -0.74
    uden
    -0.73
    aris
    -0.73
    antics
    -0.72
    ghai
    -0.71
    iquette
    -0.70
    utterstock
    -0.70
    oked
    -0.69
    POSITIVE LOGITS
     Yugoslav
    1.12
     Yugoslavia
    1.09
     Soviet
    0.99
     classmate
    0.94
     president
    0.93
     colleague
    0.91
     employee
    0.89
     President
    0.89
     comrade
    0.88
     presidents
    0.85
    Activations Density 0.551%

    No Known Activations