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. 25953
    Prev
    Next
    INDEX
    Explanations

    the phrase "without further ado."

    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
    221
    +0.11
    0.5%
    402
    +0.09
    0.4%
    472
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    221
    +0.11
    0.02
    460
    +0.09
    0.02
    455
    +0.09
    0.02
    Negative Logits
    rano
    -0.76
    orno
    -0.75
    cius
    -0.74
    ipel
    -0.70
    rolley
    -0.67
    entimes
    -0.64
    gyn
    -0.64
    inho
    -0.63
     unconsciously
    -0.63
    ldom
    -0.62
    POSITIVE LOGITS
     ado
    1.24
     developments
    0.78
     instructions
    0.78
     comment
    0.75
     inquiries
    0.73
     information
    0.73
     interruption
    0.73
     instruction
    0.73
     detail
    0.72
     explanation
    0.72
    Activations Density 0.010%

    No Known Activations