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. 2942
    Prev
    Next
    INDEX
    Explanations

    phrases related to news reporting and journalism

    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
    614
    +0.10
    0.5%
    342
    +0.09
    0.4%
    373
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    340
    +0.10
    0.04
    746
    +0.09
    0.03
    437
    +0.09
    0.04
    Negative Logits
     Examiner
    -0.77
     Yao
    -0.71
     Panda
    -0.70
     Bullets
    -0.69
     Wizard
    -0.63
     Sheep
    -0.62
     Chess
    -0.61
     Salam
    -0.61
     Barbar
    -0.60
     Sard
    -0.60
    POSITIVE LOGITS
    rogen
    1.18
    rogens
    1.11
    around
    0.76
    near
    0.71
    20439
    0.66
    âķIJ
    0.64
    romeda
    0.64
    ospons
    0.64
    rew
    0.64
    =-=-
    0.63
    Activations Density 0.591%

    No Known Activations