© 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. 22400
    Prev
    Next
    INDEX
    Explanations

    individuals seeking better deals and assistance in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered 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
    481
    +0.15
    0.7%
    121
    +0.11
    0.5%
    102
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    520
    +0.15
    0.05
    172
    +0.11
    0.04
    408
    +0.10
    0.07
    Negative Logits
     Prototype
    -0.69
    player
    -0.66
     conjecture
    -0.62
    Fan
    -0.61
    players
    -0.61
     Readers
    -0.61
     hypothesis
    -0.60
    ãĥ´ãĤ¡
    -0.60
     (@
    -0.59
     Squadron
    -0.59
    POSITIVE LOGITS
    their
    1.15
     their
    1.07
    Their
    0.92
     THEIR
    0.91
     theirs
    0.84
     they
    0.83
    They
    0.80
     themselves
    0.80
     Their
    0.78
    hower
    0.76
    Activations Density 0.967%

    No Known Activations