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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 3072
    5. 8-RES_FS3072-JB
    6. 1515
    Prev
    Next
    INDEX
    Explanations

    phrases related to reading and sharing stories or information

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs3072-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    3,072
    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
    715
    +0.12
    0.5%
    571
    +0.10
    0.5%
    26
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    246
    +0.12
    0.05
    715
    +0.10
    0.05
    755
    +0.10
    0.04
    Negative Logits
    xon
    -0.67
    ascal
    -0.65
    ño
    -0.65
    engeance
    -0.61
     towed
    -0.59
    ction
    -0.59
    idable
    -0.58
    posure
    -0.57
    aires
    -0.57
    akable
    -0.57
    POSITIVE LOGITS
     aloud
    1.51
    just
    1.23
     comprehension
    1.07
    dress
    1.06
    ied
    0.91
     excerpts
    0.91
    mitt
    0.90
    ying
    0.89
    Write
    0.88
    itatively
    0.85
    Activations Density 0.520%

    No Known Activations