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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 45992
    Prev
    Next
    INDEX
    Explanations

    references to recklessness or reckless behavior

    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
    23
    +0.14
    0.7%
    616
    +0.10
    0.5%
    232
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    744
    +0.14
    0.01
    23
    +0.10
    0.01
    705
    +0.10
    0.01
    Negative Logits
    ĻĤ
    -0.82
    HCR
    -0.74
    ļéĨĴ
    -0.70
     Rite
    -0.66
     recess
    -0.65
     Parenthood
    -0.65
    ¥µ
    -0.63
     Antar
    -0.63
     Transcript
    -0.62
     subsistence
    -0.61
    POSITIVE LOGITS
    lessly
    1.54
    lessness
    1.21
    oner
    1.18
    oning
    1.11
    ons
    1.05
    nesses
    1.02
    onen
    0.93
    entimes
    0.91
    aging
    0.90
    aged
    0.86
    Activations Density 0.005%

    No Known Activations