© 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 3072
    9. 8-RES_FS3072-JB
    11. 224
    Prev
    Next
    INDEX
    Explanations

    technical terms and acronyms related to the field of computer science

    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
    167
    +0.13
    0.6%
    551
    +0.13
    0.6%
    497
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.13
    0.12
    232
    +0.13
    0.09
    343
    +0.12
    0.09
    Negative Logits
    UGE
    -0.81
    ting
    -0.77
    fast
    -0.76
    makers
    -0.71
    erent
    -0.70
    leck
    -0.70
    hirt
    -0.68
     Rasmussen
    -0.66
    uge
    -0.66
    eren
    -0.66
    POSITIVE LOGITS
    nered
    0.93
     Rodham
    0.91
    prus
    0.90
    nea
    0.83
    bara
    0.83
    athetic
    0.80
    ologies
    0.75
    obal
    0.74
    ocaly
    0.74
    atom
    0.74
    Activations Density 1.394%

    No Known Activations