© 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 49152
    9. 8-RES_FS49152-JB
    11. 25705
    Prev
    Next
    INDEX
    Explanations

    adjectives and descriptors that indicate performance or quality

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    277
    +0.11
    0.5%
    608
    +0.09
    0.4%
    344
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    122
    +0.11
    0.04
    284
    +0.09
    0.04
    408
    +0.09
    0.04
    Negative Logits
    ulhu
    -0.88
    objects
    -0.83
    artifacts
    -0.82
     urgently
    -0.82
    Ingredients
    -0.80
    SPONSORED
    -0.79
    ifles
    -0.79
    upon
    -0.77
    types
    -0.75
    iddles
    -0.73
    POSITIVE LOGITS
     outing
    1.24
     weekend
    1.23
     upbringing
    1.17
     stint
    1.15
     offseason
    1.09
     week
    1.09
     season
    1.07
     shootout
    1.00
     month
    1.00
     summer
    0.99
    Activations Density 0.128%

    No Known Activations