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. 360
    Prev
    Next
    INDEX
    Explanations

    time-related information or references

    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
    684
    +0.11
    0.5%
    334
    +0.11
    0.5%
    591
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    651
    +0.11
    0.06
    684
    +0.11
    0.04
    132
    +0.11
    0.04
    Negative Logits
    avorite
    -0.85
    halla
    -0.68
    haar
    -0.64
     proced
    -0.61
     RAID
    -0.59
     glove
    -0.59
    mers
    -0.58
    inav
    -0.56
    itary
    -0.56
    tty
    -0.56
    POSITIVE LOGITS
    cens
    0.80
    ZI
    0.72
    ounce
    0.69
    bell
    0.66
    uncture
    0.66
    frames
    0.65
    zone
    0.64
    æĸ
    0.64
     periods
    0.62
     Unsure
    0.62
    Activations Density 0.461%

    No Known Activations