© 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. 9377
    Prev
    Next
    INDEX
    Explanations

    places or locations

    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
    420
    +0.11
    0.5%
    288
    +0.11
    0.5%
    211
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    211
    +0.11
    0.01
    515
    +0.11
    0.02
    117
    +0.10
    0.02
    Negative Logits
     Beir
    -0.73
    matter
    -0.72
    terday
    -0.71
    ×ķ
    -0.70
     Luthor
    -0.64
    urally
    -0.62
    mble
    -0.62
    ives
    -0.60
    anson
    -0.60
    iferation
    -0.60
    POSITIVE LOGITS
    onet
    1.30
    esian
    1.28
    ou
    0.97
     Area
    0.90
    onne
    0.82
    hawks
    0.81
    rina
    0.79
    liner
    0.78
    leigh
    0.77
    front
    0.76
    Activations Density 0.015%

    No Known Activations