© 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 1536
    9. 8-RES_FS1536-JB
    11. 1092
    Prev
    Next
    INDEX
    Explanations

    names of people and specific terms related to technical equipment and procedures

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    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
    377
    +0.15
    0.7%
    283
    +0.13
    0.6%
    484
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.15
    0.14
    288
    +0.13
    0.16
    343
    +0.12
    0.11
    Negative Logits
    encing
    -0.69
     Nadu
    -0.65
    encers
    -0.62
    pter
    -0.61
    uality
    -0.61
    phis
    -0.60
    mons
    -0.60
    arily
    -0.60
    apers
    -0.59
     Pradesh
    -0.59
    POSITIVE LOGITS
    strap
    0.99
    wana
    0.96
    keley
    0.89
    EGIN
    0.89
    etooth
    0.83
    hari
    0.83
    yond
    0.82
    jamin
    0.81
    oreal
    0.81
    arest
    0.81
    Activations Density 2.543%

    No Known Activations