© 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. 15998
    Prev
    Next
    INDEX
    Explanations

    information about events and activities, such as parties, movies, and projects, with a focus on providing details and instructions

    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
    497
    +0.12
    0.5%
    48
    +0.12
    0.5%
    297
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    297
    +0.12
    0.05
    546
    +0.12
    0.05
    43
    +0.11
    0.04
    Negative Logits
    terday
    -0.75
    ledged
    -0.70
    izon
    -0.66
    iday
    -0.64
    legates
    -0.64
    ļéĨĴ
    -0.61
    bidden
    -0.60
    ForgeModLoader
    -0.60
    cedented
    -0.59
    Primary
    -0.59
    POSITIVE LOGITS
     preferably
    1.45
     lest
    1.16
     please
    0.94
     eh
    0.89
     whichever
    0.85
     THEN
    0.84
     then
    0.84
    cknow
    0.84
     bask
    0.84
     beware
    0.82
    Activations Density 0.249%

    No Known Activations