Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 24576
    5. 8-RES_FS24576-JB
    6. 21311
    Prev
    Next
    INDEX
    Explanations

    locations or places within a community

    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
    731
    +0.11
    0.5%
    680
    +0.10
    0.5%
    211
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    752
    +0.11
    0.07
    731
    +0.10
    0.05
    211
    +0.10
    0.04
    Negative Logits
    thia
    -0.83
    cedes
    -0.80
    thood
    -0.79
    verbs
    -0.78
    uria
    -0.78
    vana
    -0.76
    morrow
    -0.74
    paying
    -0.74
    duty
    -0.73
    umenthal
    -0.73
    POSITIVE LOGITS
     skyline
    1.04
     Police
    0.97
     chapter
    0.95
     metropolitan
    0.94
     suburbs
    0.89
     Convention
    0.89
     Regional
    0.89
     Metropolitan
    0.89
     Institute
    0.89
     branch
    0.88
    Activations Density 0.173%

    No Known Activations