© 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. 20417
    Prev
    Next
    INDEX
    Explanations

    phrases and words related to association or connection

    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
    679
    +0.15
    0.7%
    253
    +0.12
    0.5%
    346
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    253
    +0.15
    0.02
    403
    +0.12
    0.02
    205
    +0.11
    0.02
    Negative Logits
    aneers
    -0.74
    stall
    -0.70
    tein
    -0.69
    nl
    -0.68
    umblr
    -0.66
    OUT
    -0.65
    zan
    -0.64
    ciples
    -0.64
    AIR
    -0.63
    dragon
    -0.61
    POSITIVE LOGITS
    atively
    0.97
    ively
    0.94
    ativity
    0.91
    ative
    0.82
    enza
    0.71
    atable
    0.71
     associations
    0.69
    ually
    0.68
    hips
    0.68
    eering
    0.68
    Activations Density 0.037%

    No Known Activations