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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 12288
    5. 8-RES_FS12288-JB
    6. 241
    Prev
    Next
    INDEX
    Explanations

    verbs related to examining and reviewing documents or situations

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    654
    +0.12
    0.5%
    715
    +0.11
    0.5%
    71
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    405
    +0.12
    0.05
    654
    +0.11
    0.04
    665
    +0.10
    0.05
    Negative Logits
    Torrent
    -0.74
    guard
    -0.73
    trap
    -0.70
    cil
    -0.69
    lez
    -0.69
    breaker
    -0.68
    lisher
    -0.67
    \-
    -0.66
    vic
    -0.65
    bott
    -0.64
    POSITIVE LOGITS
     favorably
    0.83
     whether
    0.82
     closely
    0.82
    ively
    0.81
     feasibility
    0.78
     extensively
    0.78
     how
    0.76
     carefully
    0.75
     trends
    0.74
    ibly
    0.72
    Activations Density 0.214%

    No Known Activations