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 49152
    5. 8-RES_FS49152-JB
    6. 41184
    Prev
    Next
    INDEX
    Explanations

    references to legislative actions and significant dates related to events or reforms

    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
    480
    +0.47
    2.4%
    627
    +0.09
    0.5%
    266
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    310
    +0.47
    0.04
    627
    +0.09
    0.03
    480
    +0.09
    0.00
    Negative Logits
     behav
    -0.63
    %.
    -0.61
    catentry
    -0.59
     thereof
    -0.59
    $.
    -0.57
     etc
    -0.56
    ifle
    -0.56
    âĵĺ
    -0.56
    izoph
    -0.55
    edi
    -0.54
    POSITIVE LOGITS
     October
    0.74
     July
    0.73
     Sept
    0.70
     August
    0.69
     September
    0.67
     April
    0.67
     December
    0.67
     February
    0.67
     March
    0.66
     November
    0.66
    Activations Density 0.323%

    No Known Activations