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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & 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. 1461
    Prev
    Next
    INDEX
    Explanations

    technology-related terms, especially those related to consumer products and devices

    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
    +0.14
    0.7%
    459
    +0.13
    0.6%
    642
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.14
    0.14
    575
    +0.13
    0.09
    459
    +0.12
    0.06
    Negative Logits
    ologies
    -0.84
    imoto
    -0.76
    ologist
    -0.74
    night
    -0.73
    ology
    -0.72
    itud
    -0.72
    enic
    -0.71
    tered
    -0.70
    orah
    -0.69
    set
    -0.69
    POSITIVE LOGITS
     actionGroup
    0.78
     goods
    0.77
     sized
    0.71
    ible
    0.69
     dur
    0.69
     electronics
    0.68
     Spending
    0.64
     Staples
    0.63
    itarian
    0.63
    erers
    0.63
    Activations Density 1.342%

    No Known Activations