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

    Neuronpedia

    Jacobian LensNEW
    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. 7758
    Prev
    Next
    INDEX
    Explanations

    terms related to data storage and retrieval, especially in the context of software development and technology

    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
    167
    +0.11
    0.5%
    216
    +0.11
    0.5%
    288
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    206
    +0.11
    0.02
    216
    +0.11
    0.02
    74
    +0.10
    0.02
    Negative Logits
    eds
    -0.79
    xxx
    -0.77
    nered
    -0.70
     Santana
    -0.70
    sky
    -0.69
    geon
    -0.63
    vous
    -0.63
    uyomi
    -0.63
    stad
    -0.63
     Cosponsors
    -0.63
    POSITIVE LOGITS
     storing
    0.91
     secrets
    0.87
    keeping
    0.79
    holders
    0.78
    abytes
    0.78
     stored
    0.77
    IER
    0.74
     shelves
    0.70
     confidential
    0.70
     indefinitely
    0.70
    Activations Density 0.046%

    No Known Activations