© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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 49152
    9. 8-RES_FS49152-JB
    11. 26782
    Prev
    Next
    INDEX
    Explanations

    time zone designations, specifically Eastern Standard Time (EST) and Central Standard Time (CST)

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    649
    +0.16
    0.7%
    323
    +0.12
    0.6%
    183
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    649
    +0.16
    0.01
    477
    +0.12
    0.01
    145
    +0.11
    0.01
    Negative Logits
    cock
    -0.76
    strap
    -0.74
    ked
    -0.72
    asive
    -0.71
    backer
    -0.70
    addons
    -0.69
    jack
    -0.67
    amac
    -0.67
    plates
    -0.66
    pton
    -0.66
    POSITIVE LOGITS
    EMBER
    0.91
    REAM
    0.80
    UTE
    0.80
    LY
    0.79
     CET
    0.78
    imation
    0.75
    orian
    0.75
    QB
    0.74
    RAL
    0.73
    OPS
    0.73
    Activations Density 0.005%

    No Known Activations