© 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. 2409
    Prev
    Next
    INDEX
    Explanations

    references to the term "PC," which likely refers to personal computers

    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
    527
    +0.13
    0.6%
    567
    +0.11
    0.5%
    99
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    527
    +0.13
    0.02
    567
    +0.11
    0.02
    417
    +0.10
    0.02
    Negative Logits
    glers
    -0.83
    atories
    -0.78
    sweet
    -0.77
    enance
    -0.76
    ammad
    -0.75
    atory
    -0.73
    terday
    -0.73
    hower
    -0.70
    ta
    -0.66
    htar
    -0.65
    POSITIVE LOGITS
     Gamer
    1.20
    ORPG
    0.99
     gamers
    0.98
     gamer
    0.95
     gaming
    0.94
    BSD
    0.93
    PDATE
    0.91
    Games
    0.90
    GS
    0.89
    Gaming
    0.87
    Activations Density 0.011%

    No Known Activations