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 1536
    5. 8-RES_FS1536-JB
    6. 1272
    Prev
    Next
    INDEX
    Explanations

    food-related terms, especially ingredients and cooking techniques

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    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
    731
    +0.10
    0.5%
    16
    +0.10
    0.4%
    464
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.10
    0.13
    469
    +0.10
    0.07
    464
    +0.10
    0.09
    Negative Logits
    adra
    -0.90
    adan
    -0.76
    icipated
    -0.71
    utra
    -0.69
    è£ħ
    -0.67
    SPONSORED
    -0.67
    owitz
    -0.66
    istries
    -0.66
    enos
    -0.65
    arat
    -0.61
    POSITIVE LOGITS
    paren
    0.93
    escription
    0.87
    dit
    0.86
    own
    0.86
    rive
    0.86
    nesday
    0.85
    river
    0.85
    eals
    0.85
    irect
    0.85
    ouble
    0.82
    Activations Density 1.968%

    No Known Activations