© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 6877
    Prev
    Next
    INDEX
    Explanations

    terms related to data collection, analysis, cybersecurity, and research funding declarations

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    519
    +0.14
    0.6%
    724
    +0.12
    0.5%
    524
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.14
    0.06
    524
    +0.12
    0.05
    724
    +0.11
    0.06
    Negative Logits
    ãĥĥãĥī
    -0.83
    urnal
    -0.79
    zl
    -0.78
    hift
    -0.73
    SN
    -0.73
    ofi
    -0.69
    otta
    -0.67
    ulet
    -0.66
    mic
    -0.65
    ãĥ´
    -0.64
    POSITIVE LOGITS
     techniques
    1.01
     procedures
    0.92
     machinery
    0.89
     prevention
    0.88
     methods
    0.87
     tools
    0.87
     chores
    0.86
     schemes
    0.85
     process
    0.84
     efforts
    0.83
    Activations Density 0.161%

    No Known Activations