Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 24576
    5. 8-RES_FS24576-JB
    6. 18952
    Prev
    Next
    INDEX
    Explanations

    personal statements or commitments

    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
    373
    +0.16
    0.7%
    604
    +0.13
    0.6%
    181
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    566
    +0.16
    0.04
    763
    +0.13
    0.03
    49
    +0.10
    0.03
    Negative Logits
    ioned
    -0.71
    Frank
    -0.66
    entimes
    -0.61
    Eastern
    -0.61
     lax
    -0.59
     airs
    -0.58
    apo
    -0.56
    APS
    -0.55
    ongevity
    -0.55
    /"
    -0.53
    POSITIVE LOGITS
     browser
    0.86
     hereby
    0.83
     Password
    0.79
     Browser
    0.74
     Login
    0.73
     yourselves
    0.73
     Javascript
    0.70
     Welcome
    0.67
    Tube
    0.66
    arest
    0.65
    Activations Density 0.146%

    No Known Activations