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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsBlogSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 49152
    5. 8-RES_FS49152-JB
    6. 43457
    Prev
    Next
    INDEX
    Explanations

    phrases related to player substitutions and absences in sports

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    447
    +0.22
    1.0%
    469
    +0.12
    0.6%
    69
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    469
    +0.22
    0.03
    222
    +0.12
    0.02
    665
    +0.10
    0.03
    Negative Logits
    sylv
    -0.81
    ipes
    -0.73
    insula
    -0.67
    orians
    -0.62
    www
    -0.60
    heres
    -0.59
    wyn
    -0.59
     splend
    -0.58
    orn
    -0.58
     Yelp
    -0.58
    POSITIVE LOGITS
     fullback
    0.87
     waivers
    0.85
     midway
    0.85
     drafted
    0.80
     Draft
    0.79
    liga
    0.78
     draft
    0.77
     captain
    0.77
     offensively
    0.76
     undrafted
    0.76
    Activations Density 0.065%

    No Known Activations