© 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. 9367
    Prev
    Next
    INDEX
    Explanations

    phrases related to turning off or cleaning up

    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
    723
    +0.10
    0.5%
    125
    +0.10
    0.5%
    593
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    125
    +0.10
    0.03
    271
    +0.10
    0.02
    262
    +0.10
    0.02
    Negative Logits
    ayer
    -0.75
    Ô
    -0.67
    ow
    -0.66
    irs
    -0.66
    iw
    -0.66
    Zero
    -0.65
    ��
    -0.64
    alf
    -0.64
    ====
    -0.62
    NetMessage
    -0.61
    POSITIVE LOGITS
     shop
    0.79
    dates
    0.73
    river
    0.72
     havoc
    0.67
    brance
    0.66
    dating
    0.63
     Budd
    0.62
    grading
    0.61
    grades
    0.61
    bath
    0.61
    Activations Density 0.054%

    No Known Activations