© 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. 22857
    Prev
    Next
    INDEX
    Explanations

    archival elements related to web content and interactions

    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
    36
    +0.29
    1.4%
    263
    +0.11
    0.5%
    372
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    36
    +0.29
    0.06
    220
    +0.11
    0.03
    250
    +0.11
    0.02
    Negative Logits
     advis
    -0.67
    mie
    -0.67
     Irving
    -0.64
    t
    -0.64
    lasses
    -0.61
     Angus
    -0.60
    ally
    -0.59
    states
    -0.59
    deck
    -0.58
     ammunition
    -0.57
    POSITIVE LOGITS
    usterity
    1.44
    esthetic
    0.98
    irst
    0.96
    cknow
    0.92
    BILITIES
    0.91
     href
    0.90
    ussie
    0.89
    ption
    0.88
    qua
    0.86
    ctors
    0.85
    Activations Density 0.044%

    No Known Activations