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. 41429
    Prev
    Next
    INDEX
    Explanations

    references to "City Hall"

    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
    447
    +0.14
    0.6%
    31
    +0.10
    0.5%
    687
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    693
    +0.14
    0.01
    31
    +0.10
    0.01
    388
    +0.09
    0.01
    Negative Logits
    hered
    -0.73
    DIT
    -0.73
    lass
    -0.72
    utsche
    -0.69
    erker
    -0.67
    unct
    -0.66
     crochet
    -0.64
    olean
    -0.64
    yip
    -0.64
    wolves
    -0.64
    POSITIVE LOGITS
     Plaza
    0.89
     Correspond
    0.84
    ibur
    0.83
    rooms
    0.83
     Buildings
    0.82
     Building
    0.81
     Auditor
    0.79
    eteria
    0.78
     offices
    0.76
    holders
    0.75
    Activations Density 0.010%

    No Known Activations