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 24576
    5. 8-RES_FS24576-JB
    6. 1154
    Prev
    Next
    INDEX
    Explanations

    violent actions or crimes

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    153
    +0.11
    0.5%
    464
    +0.10
    0.4%
    114
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.11
    0.04
    199
    +0.10
    0.03
    235
    +0.10
    0.03
    Negative Logits
    soDeliveryDate
    -0.90
    isse
    -0.72
    lopp
    -0.69
    kefeller
    -0.69
    nai
    -0.69
    Tok
    -0.67
    amous
    -0.65
    illac
    -0.65
    flex
    -0.65
     shin
    -0.64
    POSITIVE LOGITS
     manslaughter
    1.27
     offences
    0.97
     aggravated
    0.88
     charges
    0.87
     stemming
    0.84
     felony
    0.81
     misdemeanor
    0.80
     punishable
    0.77
     burglary
    0.77
    ionage
    0.77
    Activations Density 0.098%

    No Known Activations