© 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. 9424
    Prev
    Next
    INDEX
    Explanations

    information or discussions related to technology and government matters

    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
    87
    +0.46
    2.8%
    373
    +0.13
    0.8%
    447
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    87
    +0.46
    -0.01
    561
    +0.13
    0.08
    373
    +0.12
    -0.09
    Negative Logits
     UNCLASSIFIED
    -0.62
    anwhile
    -0.62
     pse
    -0.58
    ogether
    -0.55
     conflic
    -0.55
     Azerb
    -0.54
    jri
    -0.54
    _.
    -0.53
    âķIJ
    -0.53
     cryst
    -0.53
    POSITIVE LOGITS
     Belfast
    0.74
     âĢº
    0.61
     Donald
    0.59
     Transgender
    0.52
     âĢİ
    0.51
     Posted
    0.50
     Hillary
    0.50
     Calgary
    0.49
    utterstock
    0.49
     Copyright
    0.48
    Activations Density 1.875%

    No Known Activations