© 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. 17426
    Prev
    Next
    INDEX
    Explanations

    articles that emphasize specific proportions and distributions

    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
    36
    +0.15
    0.7%
    649
    +0.10
    0.4%
    752
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    36
    +0.15
    0.03
    284
    +0.10
    0.02
    210
    +0.10
    0.02
    Negative Logits
    CTV
    -0.74
    posted
    -0.73
    uden
    -0.71
    ultimate
    -0.71
    aris
    -0.69
    tsky
    -0.68
    Catal
    -0.68
    inous
    -0.67
    forth
    -0.67
    eless
    -0.67
    POSITIVE LOGITS
     representation
    0.91
     deal
    0.91
     thing
    0.89
     footprint
    0.89
     chunk
    0.88
     foothold
    0.88
     timeframe
    0.87
     swath
    0.86
     piece
    0.85
     predictor
    0.85
    Activations Density 0.072%

    No Known Activations