© 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. 2938
    Prev
    Next
    INDEX
    Explanations

    technical terms related to computational geometry and algorithms

    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
    62
    +0.12
    0.5%
    699
    +0.11
    0.5%
    651
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.12
    0.01
    23
    +0.11
    0.01
    699
    +0.11
    0.01
    Negative Logits
    ãģĦ
    -0.91
    士
    -0.85
    IDER
    -0.82
    nesota
    -0.75
     silence
    -0.72
    OAD
    -0.71
    ãģį
    -0.70
    ously
    -0.65
    Ø©
    -0.65
     Marketable
    -0.62
    POSITIVE LOGITS
    hered
    1.34
    anus
    1.31
    anic
    1.16
    oxin
    1.10
    rox
    1.08
    ris
    1.06
    rop
    1.05
    raph
    1.03
    ractor
    1.03
    rell
    1.02
    Activations Density 0.020%

    No Known Activations