© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 19922
    Prev
    Next
    INDEX
    Explanations

    sentences expressing emotional distress and self-harm ideation

    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
    665
    +0.13
    0.6%
    458
    +0.10
    0.4%
    92
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    665
    +0.13
    0.06
    195
    +0.10
    0.03
    604
    +0.09
    0.03
    Negative Logits
    aird
    -0.74
    boot
    -0.70
    usa
    -0.69
    iami
    -0.67
    enf
    -0.66
    alde
    -0.63
    luster
    -0.62
    anch
    -0.62
    AUT
    -0.61
    avery
    -0.61
    POSITIVE LOGITS
     aloud
    1.03
     oneself
    0.98
     anything
    0.96
     something
    0.91
     loudly
    0.90
    ively
    0.87
     how
    0.85
     yourself
    0.83
     specifics
    0.82
     truths
    0.82
    Activations Density 0.196%

    No Known Activations