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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 41447
    Prev
    Next
    INDEX
    Explanations

    statistics related to player performance and gameplay metrics

    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
    481
    +0.17
    0.8%
    447
    +0.13
    0.6%
    634
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    657
    +0.17
    0.03
    651
    +0.13
    0.03
    540
    +0.10
    0.03
    Negative Logits
    sonian
    -0.71
     MFT
    -0.70
    enegger
    -0.69
    onomous
    -0.67
    thouse
    -0.66
     Wik
    -0.65
    ITAL
    -0.65
     Marriott
    -0.65
    SPONSORED
    -0.64
    urtle
    -0.64
    POSITIVE LOGITS
    peed
    1.11
    pread
    1.03
     totals
    1.00
     Gained
    0.99
     accrued
    0.98
    ynthesis
    0.91
     earned
    0.89
     accumulated
    0.88
    manship
    0.86
    heet
    0.84
    Activations Density 0.070%

    No Known Activations