© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 9519
    Prev
    Next
    INDEX
    Explanations

    references to height or level of something, particularly when used in a comparative context

    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
    623
    +0.11
    0.5%
    253
    +0.09
    0.4%
    629
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    262
    +0.11
    0.03
    495
    +0.09
    0.03
    623
    +0.09
    0.02
    Negative Logits
    pty
    -0.76
     Canaver
    -0.65
    mir
    -0.64
    icia
    -0.63
    TED
    -0.63
    oris
    -0.62
    ãĤ£
    -0.62
    Deal
    -0.61
    Reply
    -0.61
    Ring
    -0.60
    POSITIVE LOGITS
     than
    1.04
     levels
    0.97
     tiers
    0.95
     priced
    0.92
    than
    0.91
     earners
    0.91
     grades
    0.88
     pitched
    0.85
     ceilings
    0.84
     priority
    0.82
    Activations Density 0.030%

    No Known Activations