© 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. 4180
    Prev
    Next
    INDEX
    Explanations

    programming code snippets that involve looping constructs

    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
    314
    +0.15
    0.7%
    373
    +0.13
    0.6%
    581
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    159
    +0.15
    0.03
    581
    +0.13
    0.03
    373
    +0.11
    -0.02
    Negative Logits
    Cho
    -0.58
    gall
    -0.57
    egu
    -0.55
    cheat
    -0.55
    yang
    -0.55
     certify
    -0.55
     Deliver
    -0.53
    REL
    -0.53
    aud
    -0.51
    floor
    -0.51
    POSITIVE LOGITS
     sake
    1.65
     purposes
    1.36
     reasons
    1.27
    ummies
    1.10
     reason
    0.94
     Reasons
    0.94
    aughs
    0.89
     ages
    0.83
     purpose
    0.79
    teness
    0.77
    Activations Density 0.156%

    No Known Activations