© 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 EvalsExportsAPI 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. 10678
    Prev
    Next
    INDEX
    Explanations

    descriptions related to physical appearance, specifically focusing on details about clothing and accessories

    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
    214
    +0.11
    0.5%
    73
    +0.10
    0.5%
    0
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    678
    +0.11
    0.02
    214
    +0.10
    0.01
    260
    +0.10
    0.02
    Negative Logits
    yrinth
    -1.01
    eer
    -0.94
    eers
    -0.89
    icative
    -0.84
    gom
    -0.84
    âĸĪâĸĪ
    -0.81
    aeus
    -0.81
    esis
    -0.80
    emort
    -0.80
    inction
    -0.79
    POSITIVE LOGITS
     blades
    1.27
     blade
    0.99
    bone
    0.94
     shoulder
    0.91
     straps
    0.90
     pads
    0.90
     shrug
    0.85
     strap
    0.85
     holster
    0.80
    chair
    0.79
    Activations Density 0.026%

    No Known Activations