© 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. 24003
    Prev
    Next
    INDEX
    Explanations

    references to something being established or confirmed as effective or valid

    oai_token-act-pair · gpt-4o-miniTriggered 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
    481
    +0.13
    0.6%
    727
    +0.09
    0.4%
    284
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    429
    +0.13
    0.02
    653
    +0.09
    0.02
    284
    +0.09
    0.02
    Negative Logits
    utterstock
    -0.84
    isons
    -0.78
    alon
    -0.78
    regon
    -0.74
    iture
    -0.74
    querade
    -0.73
    irlf
    -0.70
    regate
    -0.69
    adra
    -0.68
    hift
    -0.67
    POSITIVE LOGITS
    iary
    0.89
     proven
    0.82
     precedent
    0.78
     pedigree
    0.72
     method
    0.72
     liar
    0.68
     scientific
    0.65
     reliable
    0.65
    icity
    0.64
     causal
    0.64
    Activations Density 0.060%

    No Known Activations