© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 24000
    Prev
    Next
    INDEX
    Explanations

    terms related to stability or lack of change

    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
    288
    +0.11
    0.5%
    496
    +0.11
    0.5%
    685
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    151
    +0.11
    0.02
    288
    +0.11
    0.02
    724
    +0.10
    0.02
    Negative Logits
    udic
    -0.69
    aph
    -0.66
    omaly
    -0.66
    ueller
    -0.66
     rhet
    -0.64
    asio
    -0.63
    bull
    -0.61
    uci
    -0.60
    ammy
    -0.60
     glim
    -0.59
    POSITIVE LOGITS
    aneously
    0.84
     except
    0.82
    iated
    0.80
     unchanged
    0.80
    lihood
    0.74
    aneous
    0.74
    ledged
    0.73
    iations
    0.69
     intact
    0.66
    lishes
    0.66
    Activations Density 0.052%

    No Known Activations