© 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 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. 6193
    Prev
    Next
    INDEX
    Explanations

    words related to expressing concern or disappointment in a community or group

    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
    633
    +0.11
    0.5%
    671
    +0.09
    0.4%
    617
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    510
    +0.11
    0.03
    750
    +0.09
    0.03
    697
    +0.09
    0.03
    Negative Logits
     vetted
    -0.74
    arthed
    -0.73
    combe
    -0.73
    approved
    -0.72
     favourable
    -0.69
    artney
    -0.68
     dexter
    -0.67
    fman
    -0.67
     charism
    -0.67
    yip
    -0.67
    POSITIVE LOGITS
     :(
    0.99
     inflicted
    0.80
     ignorance
    0.79
     shame
    0.76
     injustice
    0.76
     inconvenience
    0.75
     Clown
    0.75
     unworthy
    0.74
     pity
    0.74
     betrayal
    0.74
    Activations Density 0.251%

    No Known Activations