Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsBlogSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 49152
    5. 8-RES_FS49152-JB
    6. 43462
    Prev
    Next
    INDEX
    Explanations

    terms related to artificial intelligence and deep learning

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    447
    +0.21
    1.0%
    589
    +0.13
    0.6%
    17
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    575
    +0.21
    0.02
    589
    +0.13
    0.02
    288
    +0.12
    0.02
    Negative Logits
    swick
    -0.79
    lehem
    -0.77
     Kuwait
    -0.75
     Neh
    -0.73
    fam
    -0.73
    emate
    -0.72
    iquette
    -0.72
    Þ
    -0.70
     Mog
    -0.70
     Daylight
    -0.69
    POSITIVE LOGITS
     algorithms
    1.50
     algorithm
    1.16
     technologies
    1.08
     technology
    1.03
     analytics
    1.01
     neural
    1.00
     architectures
    1.00
     techniques
    0.99
     optimization
    0.99
     processors
    0.94
    Activations Density 0.031%

    No Known Activations