© 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. 22293
    Prev
    Next
    INDEX
    Explanations

    phrases that emphasize the significance of certain outcomes or conclusions

    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
    75
    +0.13
    0.6%
    413
    +0.11
    0.5%
    131
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    75
    +0.13
    0.05
    413
    +0.11
    0.05
    242
    +0.09
    0.04
    Negative Logits
    ades
    -0.72
    agues
    -0.72
    queue
    -0.71
     Referred
    -0.65
    quer
    -0.63
     consulted
    -0.62
    ade
    -0.62
     inspected
    -0.62
    finder
    -0.62
    accompanied
    -0.61
    POSITIVE LOGITS
     how
    0.88
     desperation
    0.83
     why
    0.80
     impat
    0.79
    why
    0.77
     maturity
    0.75
     humility
    0.75
     sincerity
    0.74
     insecurity
    0.74
     emptiness
    0.73
    Activations Density 0.196%

    No Known Activations