© 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. 13081
    Prev
    Next
    INDEX
    Explanations

    the prefix "un-" followed by a verb, indicating an action that is undone or reversed

    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
    102
    +0.16
    0.8%
    616
    +0.15
    0.7%
    251
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    616
    +0.16
    0.05
    251
    +0.15
    0.05
    102
    +0.12
    0.05
    Negative Logits
     Dynamics
    -0.77
    OPLE
    -0.73
     Tut
    -0.71
     Away
    -0.71
     briefs
    -0.69
     Delivery
    -0.68
     Architecture
    -0.68
     Blitz
    -0.68
     Ajax
    -0.67
     Psychology
    -0.66
    POSITIVE LOGITS
    balanced
    1.35
    assuming
    1.30
    ruly
    1.24
    earned
    1.22
    ifying
    1.21
    availability
    1.21
    readable
    1.20
    cles
    1.20
    anticipated
    1.19
    educated
    1.18
    Activations Density 0.024%

    No Known Activations