© 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. 21994
    Prev
    Next
    INDEX
    Explanations

    words related to personal experiences and emotions, particularly focusing on heartbreak and personal challenges

    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
    122
    +0.10
    0.5%
    611
    +0.09
    0.4%
    206
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    636
    +0.10
    0.02
    711
    +0.09
    0.01
    36
    +0.09
    0.02
    Negative Logits
    esar
    -0.78
    lectic
    -0.70
    endix
    -0.68
    ffe
    -0.67
    cin
    -0.65
    lins
    -0.64
    mes
    -0.64
    unct
    -0.64
    issance
    -0.63
    eger
    -0.63
    POSITIVE LOGITS
     accountable
    1.05
     hostage
    0.97
     ransom
    0.87
     reins
    0.85
     breaths
    0.84
     breath
    0.82
     scissors
    0.79
     veto
    0.78
    hold
    0.76
     precedence
    0.75
    Activations Density 0.083%

    No Known Activations