© 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 24576
    9. 8-RES_FS24576-JB
    11. 8552
    Prev
    Next
    INDEX
    Explanations

    the word "try" with various contexts

    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
    641
    +0.10
    0.5%
    665
    +0.10
    0.4%
    202
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    70
    +0.10
    0.05
    665
    +0.10
    0.04
    641
    +0.10
    0.04
    Negative Logits
    dylib
    -0.69
    head
    -0.65
    atform
    -0.62
    edly
    -0.60
    HEAD
    -0.60
    resent
    -0.59
    effect
    -0.58
    azi
    -0.57
    hip
    -0.57
    ifa
    -0.57
    POSITIVE LOGITS
     unsuccessfully
    0.99
    nir
    0.93
     harder
    0.78
    irtual
    0.74
    anke
    0.72
    ãĥīãĥ©
    0.69
    ãĤ§
    0.69
    outs
    0.67
    bles
    0.67
    ures
    0.66
    Activations Density 0.048%

    No Known Activations