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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 5298
    Prev
    Next
    INDEX
    Explanations

    words related to technical enhancements or modifications, highlighting terms related to new technology implementations or improvements such as digital tokens, technological patents, as well as legislative reforms in intelligence-related practices

    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
    314
    +0.11
    0.5%
    19
    +0.10
    0.5%
    429
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    174
    +0.11
    0.03
    128
    +0.10
    0.03
    314
    +0.09
    0.01
    Negative Logits
     Solitaire
    -0.64
     Kills
    -0.61
     Sirius
    -0.59
    :]
    -0.59
     Daylight
    -0.59
    MEN
    -0.57
     Britann
    -0.57
    ©¶æ¥µ
    -0.56
     Bronze
    -0.56
    pak
    -0.55
    POSITIVE LOGITS
    roying
    1.03
    itored
    1.03
    tenance
    0.99
    ielding
    0.91
    vised
    0.89
    aring
    0.84
    acters
    0.83
    oldown
    0.83
    avored
    0.82
    oriented
    0.81
    Activations Density 0.086%

    No Known Activations