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. 1307
    Prev
    Next
    INDEX
    Explanations

    technical instructions in a technical context

    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
    667
    +0.12
    0.5%
    614
    +0.12
    0.5%
    388
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    667
    +0.12
    0.08
    232
    +0.12
    0.08
    284
    +0.10
    0.08
    Negative Logits
     Stranger
    -0.82
     Veronica
    -0.80
     Independence
    -0.79
     Gentleman
    -0.78
     Backup
    -0.72
     secondly
    -0.71
     independ
    -0.71
    Rated
    -0.71
    idan
    -0.70
     Takeru
    -0.69
    POSITIVE LOGITS
    ktop
    0.87
    Ħ¢
    0.82
    pes
    0.81
    ocy
    0.81
    webkit
    0.80
    ot
    0.80
    OTOS
    0.80
    oplan
    0.79
    cy
    0.79
    cyclop
    0.79
    Activations Density 1.101%

    No Known Activations