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 3072
    5. 8-RES_FS3072-JB
    6. 1921
    Prev
    Next
    INDEX
    Explanations

    information related to news events and statistics

    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
    480
    +0.40
    2.1%
    351
    +0.12
    0.6%
    7
    +0.09
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    351
    +0.40
    0.09
    480
    +0.12
    -0.06
    526
    +0.09
    -0.07
    Negative Logits
    estern
    -0.57
    reetings
    -0.55
    undrum
    -0.50
    resa
    -0.49
     Anthropology
    -0.48
    eva
    -0.48
     Fasc
    -0.47
    Merit
    -0.47
     Posted
    -0.47
    ukong
    -0.47
    POSITIVE LOGITS
     accordingly
    0.85
    ).[
    0.84
     thereafter
    0.78
     afterwards
    0.75
     thereto
    0.74
     nonetheless
    0.72
     theirs
    0.68
     therein
    0.68
     afterward
    0.67
     thereof
    0.67
    Activations Density 7.529%

    No Known Activations