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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 24576
    5. 8-RES_FS24576-JB
    6. 11262
    Prev
    Next
    INDEX
    Explanations

    references to changes or comparisons

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    218
    +0.13
    0.6%
    507
    +0.10
    0.4%
    43
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    351
    +0.13
    0.03
    570
    +0.10
    0.01
    600
    +0.09
    0.03
    Negative Logits
    ortment
    -0.65
    ccording
    -0.63
     maneu
    -0.61
     periodically
    -0.60
    itton
    -0.60
     towed
    -0.59
    hers
    -0.59
     assorted
    -0.58
    leted
    -0.58
    isha
    -0.57
    POSITIVE LOGITS
     anymore
    1.55
     nor
    1.25
     anything
    1.02
     whatsoever
    1.01
    nor
    0.97
     any
    0.97
     slightest
    0.92
     anybody
    0.87
     necessarily
    0.85
     ANY
    0.82
    Activations Density 0.276%

    No Known Activations