© 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. 22783
    Prev
    Next
    INDEX
    Explanations

    mentions of governance and related administrative concepts

    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
    314
    +0.12
    0.6%
    240
    +0.12
    0.5%
    149
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.12
    0.02
    288
    +0.12
    0.03
    343
    +0.11
    0.02
    Negative Logits
    ity
    -0.64
     drill
    -0.64
     DISTR
    -0.62
    FE
    -0.62
    itous
    -0.60
     circuit
    -0.59
     WATCHED
    -0.58
     prescribe
    -0.58
     rendition
    -0.57
     Sharif
    -0.55
    POSITIVE LOGITS
    ors
    1.45
    orship
    1.44
    or
    1.16
    ements
    1.16
    ored
    1.12
    orically
    1.05
    tons
    1.04
    oric
    0.99
    orical
    0.98
    oring
    0.98
    Activations Density 0.067%

    No Known Activations