© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 12288
    9. 8-RES_FS12288-JB
    11. 10254
    Prev
    Next
    INDEX
    Explanations

    scientific terms and technical jargon related to chemistry and physics

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    174
    +0.11
    0.5%
    288
    +0.10
    0.5%
    659
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.11
    0.05
    232
    +0.10
    0.03
    284
    +0.10
    0.03
    Negative Logits
    hovah
    -0.70
    grades
    -0.70
    suit
    -0.68
    ongyang
    -0.68
    Trend
    -0.65
    yer
    -0.62
    OTUS
    -0.62
    yre
    -0.61
    ractive
    -0.61
    sonian
    -0.60
    POSITIVE LOGITS
    igans
    1.41
    ufact
    1.28
    omics
    1.13
    thal
    1.05
    igan
    1.02
    ovich
    1.01
    atural
    1.00
    ovic
    0.98
    ovember
    0.97
    ews
    0.96
    Activations Density 0.180%

    No Known Activations