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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsBlogSlackPrivacy & 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. 37057
    Prev
    Next
    INDEX
    Explanations

    concepts related to societal issues and challenges, particularly those involving injustice, rights, and labor

    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
    481
    +0.20
    0.9%
    447
    +0.17
    0.8%
    724
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.20
    0.07
    724
    +0.17
    0.07
    225
    +0.10
    0.06
    Negative Logits
    ãĥ´
    -0.61
    ãĤ±
    -0.59
    APTER
    -0.57
    ificant
    -0.57
    inosaur
    -0.57
    Mos
    -0.56
    ghazi
    -0.56
    itialized
    -0.55
    ãĤ¨ãĥ«
    -0.54
    MAC
    -0.53
    POSITIVE LOGITS
     and
    0.78
     syndrome
    0.77
     or
    0.75
     rather
    0.72
     techniques
    0.72
    .
    0.72
     disorder
    0.71
     ("
    0.69
     tactics
    0.67
     therapy
    0.66
    Activations Density 0.272%

    No Known Activations