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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsBlogSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 768
    5. 8-RES_FS768-JB
    6. 289
    Prev
    Next
    INDEX
    Explanations

    locations or institutions related to educational or legal matters

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs768-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    768
    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
    659
    +0.14
    0.6%
    232
    +0.11
    0.5%
    24
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.14
    0.24
    232
    +0.11
    0.21
    284
    +0.11
    0.17
    Negative Logits
     fictitious
    -0.71
     ster
    -0.71
     scra
    -0.70
     compan
    -0.69
     cryst
    -0.68
     fec
    -0.66
     reflective
    -0.66
    PDATE
    -0.65
     resil
    -0.65
     reflection
    -0.64
    POSITIVE LOGITS
    oglu
    1.21
    aja
    1.12
    icz
    1.11
    ka
    1.10
    ji
    1.09
    sky
    1.07
    án
    1.07
    owski
    1.07
    akis
    1.06
    nik
    1.06
    Activations Density 3.456%

    No Known Activations