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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & 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. 43392
    Prev
    Next
    INDEX
    Explanations

    phrases related to the nature and concept of existence

    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
    +0.15
    0.7%
    7
    +0.15
    0.7%
    255
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    102
    +0.15
    0.07
    7
    +0.15
    0.06
    512
    +0.12
    0.05
    Negative Logits
    çīĪ
    -0.91
    luaj
    -0.81
    osate
    -0.78
    itches
    -0.78
     throats
    -0.77
    IFA
    -0.75
    ahs
    -0.74
    ortmund
    -0.71
    ynasty
    -0.71
    backer
    -0.71
    POSITIVE LOGITS
     subjective
    1.22
     inherently
    1.18
     intrinsically
    1.16
     finite
    1.03
     insepar
    1.02
     immutable
    1.01
     fleeting
    0.99
     ALWAYS
    0.99
     foundational
    0.97
     meaningless
    0.97
    Activations Density 0.189%

    No Known Activations