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 3072
    5. 8-RES_FS3072-JB
    6. 2440
    Prev
    Next
    INDEX
    Explanations

    economic and financial terms and concepts

    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
    314
    +0.13
    0.6%
    222
    +0.11
    0.5%
    673
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.13
    0.13
    575
    +0.11
    0.10
    222
    +0.11
    0.05
    Negative Logits
    ussions
    -0.75
    ones
    -0.68
     oneself
    -0.64
     Hond
    -0.64
    ILE
    -0.63
     Maurit
    -0.62
    someone
    -0.59
    something
    -0.59
     Highlander
    -0.59
    rogens
    -0.59
    POSITIVE LOGITS
    eworld
    0.79
    mma
    0.79
    ĸļ
    0.70
    âĺ
    0.69
     woes
    0.69
     selves
    0.68
    iour
    0.68
     destiny
    0.66
     empire
    0.66
    stretched
    0.66
    Activations Density 2.175%

    No Known Activations