© 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. 10027
    Prev
    Next
    INDEX
    Explanations

    phrases related to opinions and beliefs

    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
    449
    +0.12
    0.5%
    510
    +0.10
    0.5%
    105
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    449
    +0.12
    0.03
    510
    +0.10
    0.03
    675
    +0.09
    0.03
    Negative Logits
    hs
    -0.69
    Rated
    -0.69
    jab
    -0.68
    md
    -0.67
    verting
    -0.67
    recorded
    -0.67
    ithing
    -0.66
    sing
    -0.65
    ping
    -0.63
    unker
    -0.61
    POSITIVE LOGITS
    enance
    0.79
     logically
    0.79
     Esports
    0.70
     forgiven
    0.66
     Santos
    0.64
    DragonMagazine
    0.64
    OULD
    0.63
     Carbuncle
    0.63
    gat
    0.63
     ---------
    0.61
    Activations Density 0.080%

    No Known Activations