© 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. 287
    Prev
    Next
    INDEX
    Explanations

    Phrases related to debates or choices, particularly centered around conflicting views or decisions

    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
    456
    +0.13
    0.6%
    236
    +0.10
    0.5%
    255
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    456
    +0.13
    0.04
    236
    +0.10
    0.04
    60
    +0.10
    0.03
    Negative Logits
    Catalog
    -0.95
    Eye
    -0.77
    lator
    -0.73
    oven
    -0.73
    oxide
    -0.70
    oola
    -0.69
    icker
    -0.69
    agraph
    -0.69
    venth
    -0.68
    wich
    -0.68
    POSITIVE LOGITS
    soever
    0.93
     they
    0.82
     anyone
    0.78
     there
    0.75
     respondents
    0.71
     anybody
    0.69
     it
    0.67
     he
    0.67
     intentional
    0.65
     consciously
    0.65
    Activations Density 0.037%

    No Known Activations