© 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. 1990
    Prev
    Next
    INDEX
    Explanations

    Over-represent the word "Over" in the text

    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
    269
    +0.11
    0.5%
    640
    +0.10
    0.4%
    365
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    269
    +0.11
    0.03
    684
    +0.10
    0.03
    355
    +0.10
    0.03
    Negative Logits
    BILITY
    -0.65
    osphere
    -0.60
     Nirvana
    -0.59
     Phantom
    -0.57
    arium
    -0.57
    ä½ľ
    -0.57
     understatement
    -0.56
    ä
    -0.56
    BLE
    -0.55
     Dot
    -0.55
    POSITIVE LOGITS
    whelming
    1.19
    whel
    1.12
    lap
    1.08
    flow
    1.00
    heard
    1.00
    dose
    0.99
    ride
    0.99
    haul
    0.99
    statement
    0.99
    arching
    0.99
    Activations Density 0.035%

    No Known Activations