© 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. 13293
    Prev
    Next
    INDEX
    Explanations

    short descriptions or accounts

    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
    194
    +0.11
    0.5%
    222
    +0.11
    0.5%
    725
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    194
    +0.11
    0.03
    222
    +0.11
    0.03
    402
    +0.10
    0.03
    Negative Logits
    ILLE
    -0.78
    IRO
    -0.75
    ONSORED
    -0.74
    ADRA
    -0.74
    Materials
    -0.68
    ICAN
    -0.67
    CLASSIFIED
    -0.67
     Compass
    -0.67
    Magikarp
    -0.66
    çļ
    -0.63
    POSITIVE LOGITS
    ening
    1.18
    sighted
    1.13
    ener
    1.10
    lived
    1.07
    ened
    1.05
    bread
    1.00
    comings
    0.99
    coming
    0.98
    eners
    0.94
    listed
    0.93
    Activations Density 0.025%

    No Known Activations