© 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. 7337
    Prev
    Next
    INDEX
    Explanations

    information related to shields and shielding

    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
    288
    +0.18
    0.8%
    238
    +0.13
    0.6%
    189
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.18
    0.02
    189
    +0.13
    0.01
    238
    +0.11
    0.01
    Negative Logits
    gres
    -0.79
    iasco
    -0.78
    pheus
    -0.72
    ASY
    -0.71
    lihood
    -0.70
     Judicial
    -0.69
     Helpful
    -0.68
     âĸº
    -0.67
    ĸļ
    -0.67
    jury
    -0.67
    POSITIVE LOGITS
    heed
    0.86
     shield
    0.84
     shields
    0.81
     shielding
    0.78
     piercing
    0.75
     absorbing
    0.75
     shielded
    0.72
     protecting
    0.72
    lain
    0.72
     wall
    0.71
    Activations Density 0.032%

    No Known Activations