© 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. 13774
    Prev
    Next
    INDEX
    Explanations

    references to temporary or provisional positions or circumstances

    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
    288
    +0.15
    0.7%
    411
    +0.09
    0.4%
    572
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.15
    0.01
    411
    +0.09
    0.01
    691
    +0.09
    0.01
    Negative Logits
    gling
    -0.86
    kie
    -0.80
    uria
    -0.79
    ongs
    -0.78
    plex
    -0.78
    atche
    -0.77
    asus
    -0.76
    ongyang
    -0.76
    wagen
    -0.76
    cius
    -0.75
    POSITIVE LOGITS
     interim
    0.88
     president
    0.82
     CEO
    0.78
     chairman
    0.75
     truce
    0.73
     leader
    0.73
     manager
    0.73
     superintendent
    0.72
     commander
    0.71
     director
    0.70
    Activations Density 0.018%

    No Known Activations