© 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 EvalsExportsinterp-engineNEWAPI 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. 16283
    Prev
    Next
    INDEX
    Explanations

    information about upcoming product releases and updates

    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
    640
    +0.33
    1.6%
    49
    +0.10
    0.5%
    443
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.33
    0.14
    94
    +0.10
    0.09
    64
    +0.09
    -0.03
    Negative Logits
    boro
    -0.74
     photographed
    -0.72
     bicycl
    -0.68
     volunteered
    -0.68
     crosses
    -0.68
    hower
    -0.67
     Stew
    -0.65
     photograp
    -0.64
     prosec
    -0.64
    yll
    -0.63
    POSITIVE LOGITS
    Samsung
    1.21
    Sony
    1.14
    Speaking
    1.12
    Xbox
    1.09
    Android
    1.07
    Nintendo
    1.03
     Razer
    1.01
    LG
    0.98
    Exit
    0.98
    Developer
    0.98
    Activations Density 0.375%

    No Known Activations