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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 49152
    5. 8-RES_FS49152-JB
    6. 41449
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "step"

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    202
    +0.11
    0.5%
    359
    +0.09
    0.4%
    74
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    202
    +0.11
    0.02
    435
    +0.09
    0.02
    13
    +0.09
    0.02
    Negative Logits
     Unic
    -0.76
    orem
    -0.75
     Pengu
    -0.74
    ortunately
    -0.72
    ILLE
    -0.69
     tiss
    -0.69
    eatures
    -0.69
    ominated
    -0.67
    essage
    -0.65
    inately
    -0.64
    POSITIVE LOGITS
    daughter
    1.18
    dad
    1.11
    brother
    1.04
    mother
    0.98
    father
    0.98
    hens
    0.89
    isters
    0.88
    steps
    0.88
    mom
    0.84
    step
    0.83
    Activations Density 0.017%

    No Known Activations