© 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. 1610
    Prev
    Next
    INDEX
    Explanations

    information related to regulations, specifically regarding products and advertising

    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
    161
    +0.11
    0.5%
    766
    +0.10
    0.5%
    48
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    194
    +0.11
    0.04
    49
    +0.10
    0.05
    718
    +0.10
    0.05
    Negative Logits
    ffe
    -0.77
    atl
    -0.72
     Maze
    -0.69
    erity
    -0.67
    forts
    -0.67
    roads
    -0.64
    miah
    -0.64
     logistics
    -0.61
     intertw
    -0.61
    uckle
    -0.61
    POSITIVE LOGITS
     disqual
    1.19
     qualifies
    1.10
     qualify
    1.04
     eligible
    0.96
     ineligible
    0.95
     disqualified
    0.92
     exempt
    0.92
     acceptable
    0.90
     deserving
    0.90
     invalid
    0.87
    Activations Density 0.692%

    No Known Activations