© 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. 21008
    Prev
    Next
    INDEX
    Explanations

    positive reviews about products or items

    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
    447
    +0.24
    1.1%
    481
    +0.13
    0.6%
    326
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.24
    0.06
    288
    +0.13
    0.06
    678
    +0.10
    0.05
    Negative Logits
    76561
    -0.76
    REAM
    -0.74
    erate
    -0.72
    judicial
    -0.71
    rians
    -0.70
    lished
    -0.66
    angering
    -0.64
     advers
    -0.63
     slowdown
    -0.63
    killing
    -0.63
    POSITIVE LOGITS
     holder
    0.93
     kit
    0.93
     pictured
    0.92
     accessories
    0.92
     accessory
    0.91
     kits
    0.90
     adorned
    0.89
     worn
    0.89
    washer
    0.87
     attachments
    0.86
    Activations Density 0.303%

    No Known Activations