© 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. 26832
    Prev
    Next
    INDEX
    Explanations

    references to hauntings and the paranormal

    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
    326
    +0.16
    0.7%
    549
    +0.14
    0.6%
    32
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.16
    0.01
    549
    +0.14
    0.01
    403
    +0.11
    0.01
    Negative Logits
    DN
    -0.79
    enture
    -0.78
    iyah
    -0.77
    cellence
    -0.75
    ivot
    -0.74
    nice
    -0.73
    Effective
    -0.73
    ettel
    -0.73
    ibaba
    -0.73
    iciency
    -0.71
    POSITIVE LOGITS
    crow
    1.05
     haunting
    1.03
     haunted
    1.00
     haunt
    1.00
     haun
    0.98
     Mansion
    0.93
    Haunted
    0.86
     Haunted
    0.84
     attractions
    0.79
     tales
    0.77
    Activations Density 0.010%

    No Known Activations