© 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. 17678
    Prev
    Next
    INDEX
    Explanations

    information related to different individuals like students, freshmen, and agents within academic and professional contexts

    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
    373
    +0.10
    0.5%
    288
    +0.10
    0.5%
    327
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    711
    +0.10
    0.03
    575
    +0.10
    0.05
    288
    +0.09
    0.05
    Negative Logits
    ologies
    -0.84
    ansk
    -0.80
    zees
    -0.79
    acists
    -0.78
     domains
    -0.76
     Rooms
    -0.74
    brids
    -0.73
    ouls
    -0.72
    onies
    -0.72
    uckles
    -0.71
    POSITIVE LOGITS
     fundraiser
    0.83
    digy
    0.81
     contributor
    0.79
     scout
    0.79
     unto
    0.79
     contestant
    0.78
     fixture
    0.76
     participant
    0.75
     contender
    0.75
     holder
    0.74
    Activations Density 0.206%

    No Known Activations