© 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. 10592
    Prev
    Next
    INDEX
    Explanations

    references to specific persons or companies

    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
    676
    +0.10
    0.5%
    415
    +0.09
    0.4%
    247
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    76
    +0.10
    0.05
    676
    +0.09
    0.05
    693
    +0.09
    0.04
    Negative Logits
    Reply
    -0.68
    thia
    -0.62
    hemy
    -0.61
     Invaders
    -0.60
    STEM
    -0.59
    ptoms
    -0.58
     Chronicles
    -0.58
    ratulations
    -0.57
    evidence
    -0.57
     Guards
    -0.56
    POSITIVE LOGITS
     behest
    1.42
     expense
    1.30
     discretion
    1.00
     level
    1.00
     intersections
    1.00
     outset
    0.96
     intervals
    0.95
     disposal
    0.95
     doorstep
    0.94
     fingertips
    0.91
    Activations Density 0.306%

    No Known Activations