© 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. 20115
    Prev
    Next
    INDEX
    Explanations

    words related to physical actions or descriptions of events, such as "face," "close quarters," "closed," "warm," and "gun."

    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
    +0.21
    1.0%
    545
    +0.11
    0.5%
    332
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    251
    +0.21
    0.03
    96
    +0.11
    0.03
    677
    +0.11
    0.03
    Negative Logits
    naire
    -0.65
     Splash
    -0.60
     Seal
    -0.60
     Duchess
    -0.59
    ressor
    -0.59
     Rolls
    -0.58
    artment
    -0.58
     unsolved
    -0.57
     Supplement
    -0.57
     leak
    -0.56
    POSITIVE LOGITS
     intervals
    0.84
     proportions
    0.82
    sight
    0.76
    omever
    0.71
     angles
    0.66
    practice
    0.66
    reads
    0.65
     doses
    0.65
    livion
    0.65
    evidence
    0.64
    Activations Density 0.244%

    No Known Activations