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. 41389
    Prev
    Next
    INDEX
    Explanations

    special characters or accented letters

    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
    373
    +0.13
    0.6%
    466
    +0.10
    0.5%
    423
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.13
    0.03
    466
    +0.10
    0.03
    107
    +0.10
    0.02
    Negative Logits
     Clover
    -0.62
     endowed
    -0.59
    inition
    -0.57
     sterling
    -0.57
     maintaining
    -0.57
    atform
    -0.56
     trumpet
    -0.56
    enegger
    -0.55
     scrolling
    -0.55
    ocaust
    -0.55
    POSITIVE LOGITS
    pta
    0.82
    ivas
    0.82
    anka
    0.80
    adic
    0.80
    oslav
    0.77
    oku
    0.75
    uner
    0.74
    atoon
    0.73
    inx
    0.73
    iso
    0.71
    Activations Density 0.038%

    No Known Activations