© 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. 20458
    Prev
    Next
    INDEX
    Explanations

    social media sharing buttons

    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
    373
    +0.10
    0.5%
    570
    +0.08
    0.4%
    366
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    257
    +0.10
    0.02
    373
    +0.08
    -0.01
    241
    +0.08
    0.02
    Negative Logits
    :]
    -0.64
    actively
    -0.58
     pacing
    -0.54
    hol
    -0.54
     tent
    -0.51
    externalActionCode
    -0.50
    quartered
    -0.48
    iba
    -0.48
     elevation
    -0.47
    collar
    -0.47
    POSITIVE LOGITS
    ificate
    0.67
    legram
    0.62
    agascar
    0.61
    estine
    0.61
    BSD
    0.60
    Tumblr
    0.60
     alike
    0.59
    romeda
    0.59
    Facebook
    0.58
    Subscribe
    0.55
    Activations Density 0.032%

    No Known Activations