© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 3736
    Prev
    Next
    INDEX
    Explanations

    instances where things overlap or are combined, especially in the context of music composition or design layers

    oai_token-act-pair · gpt-3.5-turbo
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    382
    +0.10
    0.5%
    154
    +0.10
    0.5%
    172
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.10
    0.02
    588
    +0.10
    0.01
    382
    +0.10
    0.01
    Negative Logits
    ELY
    -0.76
    ger
    -0.76
    uay
    -0.72
    =-=-=-=-=-=-=-=-
    -0.72
    Planet
    -0.72
    ified
    -0.71
    boat
    -0.71
    ista
    -0.70
    rique
    -0.68
    ifier
    -0.67
    POSITIVE LOGITS
     layers
    0.79
    lay
    0.74
    stretched
    0.69
    apped
    0.69
     overl
    0.67
    heads
    0.66
     populated
    0.65
     offending
    0.63
     borders
    0.62
     timelines
    0.61
    Activations Density 0.036%

    No Known Activations