© 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 12288
    9. 8-RES_FS12288-JB
    11. 1259
    Prev
    Next
    INDEX
    Explanations

    references to salary cap limits in sports leagues

    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
    552
    +0.13
    0.6%
    421
    +0.11
    0.5%
    238
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    552
    +0.13
    0.03
    421
    +0.11
    0.02
    316
    +0.10
    0.02
    Negative Logits
    hower
    -0.83
    ĪĴ
    -0.77
    cause
    -0.74
    perse
    -0.71
     anten
    -0.71
     silence
    -0.69
    ¿½
    -0.67
    selves
    -0.65
    isance
    -0.62
     Nare
    -0.61
    POSITIVE LOGITS
    itol
    1.52
    uchin
    1.19
    illary
    1.16
    rice
    1.14
    aic
    1.10
    itals
    1.09
    rices
    1.06
    acity
    0.98
    itated
    0.96
    ulet
    0.96
    Activations Density 0.069%

    No Known Activations