© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 2107
    Prev
    Next
    INDEX
    Explanations

    terms related to software development and coding procedures

    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
    544
    +0.11
    0.5%
    417
    +0.11
    0.5%
    557
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    417
    +0.11
    0.04
    575
    +0.11
    0.05
    25
    +0.10
    0.03
    Negative Logits
    estern
    -0.65
    soon
    -0.60
     actionGroup
    -0.58
    almost
    -0.57
     Pathfinder
    -0.54
     ''
    -0.52
    irie
    -0.52
     Catch
    -0.51
    oso
    -0.51
    hot
    -0.50
    POSITIVE LOGITS
     nor
    0.98
     anymore
    0.97
     ones
    0.91
     altogether
    0.81
     itself
    0.79
     necessarily
    0.74
     thereof
    0.74
    "!
    0.69
    %.
    0.68
     attRot
    0.67
    Activations Density 0.493%

    No Known Activations