© 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 EvalsExportsinterp-engineNEWAPI 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. 4899
    Prev
    Next
    INDEX
    Explanations

    references to mistreatment or mistreatment-related scenarios

    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
    288
    +0.15
    0.7%
    459
    +0.12
    0.5%
    19
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    52
    +0.15
    0.01
    19
    +0.12
    0.01
    289
    +0.11
    0.01
    Negative Logits
    phrine
    -0.68
     thumbs
    -0.65
    Reviewer
    -0.64
    adesh
    -0.63
     scratch
    -0.63
    */(
    -0.62
    ggies
    -0.61
     Corpus
    -0.61
    jriwal
    -0.60
    ITAL
    -0.60
    POSITIVE LOGITS
    reatment
    1.52
    reated
    1.33
    rans
    1.17
    ypes
    1.15
    reating
    1.14
    arget
    1.11
    reat
    1.01
    mallow
    1.00
    aken
    0.97
    resses
    0.96
    Activations Density 0.013%

    No Known Activations