© 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 12288
    9. 8-RES_FS12288-JB
    11. 10632
    Prev
    Next
    INDEX
    Explanations

    words related to astronomy and astrophysics

    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
    549
    +0.15
    0.7%
    288
    +0.13
    0.6%
    441
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    441
    +0.15
    0.02
    549
    +0.13
    0.02
    288
    +0.13
    0.02
    Negative Logits
     Hearts
    -0.68
    ipal
    -0.68
    ãģĭ
    -0.66
     Nightmares
    -0.65
    hood
    -0.63
     Advertisement
    -0.61
     TIME
    -0.61
    sburgh
    -0.61
    Lay
    -0.60
    士
    -0.60
    POSITIVE LOGITS
    roph
    1.28
    rology
    1.18
    igmat
    1.16
    rament
    1.15
    rolog
    1.14
    rob
    1.08
    rog
    1.07
    ral
    1.07
    rop
    1.07
    ron
    1.03
    Activations Density 0.050%

    No Known Activations