© 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 12288
    9. 8-RES_FS12288-JB
    11. 9363
    Prev
    Next
    INDEX
    Explanations

    measurements related to physical dimensions such as volume, depth, length, height, width, size, position, and opacity

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    99
    +0.16
    0.7%
    255
    +0.12
    0.5%
    35
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    255
    +0.16
    0.04
    99
    +0.12
    0.04
    678
    +0.11
    0.03
    Negative Logits
    bara
    -0.79
    JECT
    -0.68
    ctuary
    -0.66
    istine
    -0.65
    uterte
    -0.63
    andum
    -0.62
    uthor
    -0.62
    zsche
    -0.62
    dinand
    -0.61
    eer
    -0.60
    POSITIVE LOGITS
    Chart
    0.81
     constraints
    0.79
     difference
    0.78
     measurement
    0.77
     mismatch
    0.76
     Discrimination
    0.76
     Width
    0.75
     elevation
    0.74
     gradient
    0.74
     height
    0.73
    Activations Density 0.148%

    No Known Activations