© 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. 22423
    Prev
    Next
    INDEX
    Explanations

    themes of change, inner struggle, and personal transformation

    oai_token-act-pair · gpt-4o-miniTriggered 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
    640
    +0.25
    1.2%
    314
    +0.17
    0.8%
    447
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.25
    0.08
    143
    +0.17
    0.04
    6
    +0.13
    0.04
    Negative Logits
     enemy
    -0.66
     destro
    -0.64
     princ
    -0.64
    tons
    -0.61
     charm
    -0.61
     pro
    -0.60
     adversary
    -0.60
     targets
    -0.60
     backbone
    -0.59
     ACTIONS
    -0.59
    POSITIVE LOGITS
    ³³³
    0.88
    Newsletter
    0.84
    Posted
    0.84
    Synopsis
    0.82
    è¦ļéĨĴ
    0.78
    Narr
    0.77
    Introduced
    0.76
    ³³³³³³³³
    0.74
    Rating
    0.73
    posted
    0.72
    Activations Density 0.169%

    No Known Activations