© 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. 10245
    Prev
    Next
    INDEX
    Explanations

    salutations or titles referring to people

    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
    430
    +0.14
    0.6%
    176
    +0.13
    0.6%
    314
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    430
    +0.14
    0.06
    232
    +0.13
    0.04
    176
    +0.10
    0.04
    Negative Logits
     isolate
    -0.75
     destro
    -0.73
     routed
    -0.69
     actionGroup
    -0.69
     divided
    -0.66
     damp
    -0.65
     contrasting
    -0.62
     appre
    -0.62
     cruc
    -0.62
     overshadow
    -0.61
    POSITIVE LOGITS
     Universe
    0.93
     Claus
    0.93
     Robot
    0.84
     Hyde
    0.81
     Chips
    0.77
    .,
    0.76
     Spock
    0.75
     GOD
    0.75
    rahim
    0.73
     Incredible
    0.73
    Activations Density 0.061%

    No Known Activations