© 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. 7290
    Prev
    Next
    INDEX
    Explanations

    reflection on personal thoughts and decisions

    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
    405
    +0.10
    0.5%
    270
    +0.10
    0.5%
    595
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    155
    +0.10
    0.03
    405
    +0.10
    0.03
    582
    +0.10
    0.02
    Negative Logits
    owed
    -0.71
    ante
    -0.70
    ©¶æ¥µ
    -0.70
    soever
    -0.69
    owe
    -0.69
    mouth
    -0.69
    WF
    -0.69
    SUP
    -0.68
    zynski
    -0.68
    quet
    -0.67
    POSITIVE LOGITS
     how
    1.02
     nostalg
    0.90
     hypot
    0.86
     whether
    0.84
     quitting
    0.83
     ways
    0.80
     blowing
    0.79
     why
    0.78
     what
    0.78
     aloud
    0.74
    Activations Density 0.063%

    No Known Activations