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. 2540
    Prev
    Next
    INDEX
    Explanations

    historical references or mentions

    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
    288
    +0.13
    0.6%
    57
    +0.10
    0.5%
    290
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.13
    0.06
    724
    +0.10
    0.06
    72
    +0.10
    0.04
    Negative Logits
    cheon
    -0.70
    enger
    -0.69
    igans
    -0.69
    inki
    -0.69
    achment
    -0.68
    ricular
    -0.66
    aton
    -0.66
    uana
    -0.64
    aging
    -0.64
    tan
    -0.63
    POSITIVE LOGITS
    orians
    0.92
     buffs
    0.91
    arily
    0.80
     lesson
    0.78
     repeats
    0.77
     history
    0.76
     precedent
    0.76
    orical
    0.76
     nostalg
    0.74
     relics
    0.74
    Activations Density 0.597%

    No Known Activations