© 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. 7065
    Prev
    Next
    INDEX
    Explanations

    the planet names Mars, Jupiter, or Uranus

    oai_token-act-pair · gpt-3.5-turboTriggered 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
    481
    +0.12
    0.5%
    354
    +0.10
    0.5%
    695
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    47
    +0.12
    0.05
    639
    +0.10
    0.05
    278
    +0.09
    0.04
    Negative Logits
    eval
    -0.70
    ires
    -0.67
    edit
    -0.65
     Wast
    -0.60
     understandably
    -0.58
    rules
    -0.57
     props
    -0.56
    ETS
    -0.56
    Words
    -0.56
    alloc
    -0.56
    POSITIVE LOGITS
    acle
    1.44
    chard
    1.39
    acles
    1.36
    chid
    1.35
    nam
    1.33
    ifice
    1.26
    Else
    1.22
    gy
    1.21
     otherwise
    1.15
    ific
    1.11
    Activations Density 0.142%

    No Known Activations