© 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. 37540
    Prev
    Next
    INDEX
    Explanations

    actions related to decision-making and intentions

    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
    481
    +0.19
    0.9%
    186
    +0.12
    0.6%
    302
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    49
    +0.19
    0.08
    665
    +0.12
    0.07
    718
    +0.11
    0.07
    Negative Logits
     Canaveral
    -0.74
    ELD
    -0.65
    uably
    -0.64
    quickShipAvailable
    -0.60
    interstitial
    -0.60
    auga
    -0.60
    afety
    -0.60
    hua
    -0.60
     Associated
    -0.59
    senal
    -0.59
    POSITIVE LOGITS
     intervene
    1.06
     revisit
    1.03
     reconsider
    1.01
     investigate
    1.01
     remedy
    1.00
     reproduce
    0.97
     follow
    0.97
     try
    0.97
     proceed
    0.95
     pursue
    0.95
    Activations Density 0.205%

    No Known Activations