© 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. 26673
    Prev
    Next
    INDEX
    Explanations

    references to media and entertainment, particularly in relation to musical bands and related events

    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
    481
    +0.47
    2.4%
    447
    +0.24
    1.2%
    205
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    308
    +0.47
    0.04
    481
    +0.24
    0.07
    627
    +0.08
    0.03
    Negative Logits
     inputs
    -0.62
     commercials
    -0.62
     grazing
    -0.62
     nesting
    -0.61
     collateral
    -0.60
     noting
    -0.59
     coughing
    -0.58
     misdem
    -0.58
     joking
    -0.58
     aggrav
    -0.58
    POSITIVE LOGITS
    Gra
    0.63
    STER
    0.60
     Hort
    0.60
    eki
    0.60
    jad
    0.59
     Valiant
    0.59
    Welcome
    0.59
     Harvest
    0.58
    True
    0.57
     Nom
    0.57
    Activations Density 0.573%

    No Known Activations