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 6144
    5. 8-RES_FS6144-JB
    6. 4577
    Prev
    Next
    INDEX
    Explanations

    phrases related to hate and hate speech

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs6144-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_6144
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    6,144
    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
    289
    +0.13
    0.6%
    161
    +0.12
    0.6%
    420
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    161
    +0.13
    0.03
    288
    +0.12
    0.04
    32
    +0.12
    0.04
    Negative Logits
    aver
    -0.82
    agers
    -0.74
     Pione
    -0.73
    UNCH
    -0.72
    aqu
    -0.70
    enture
    -0.70
    clinton
    -0.69
     Cancel
    -0.68
    å§«
    -0.68
    inarily
    -0.68
    POSITIVE LOGITS
     hate
    0.90
     crimes
    0.89
    fully
    0.88
     mobs
    0.82
     crime
    0.81
     prejudice
    0.81
    fulness
    0.80
     hatred
    0.80
     vengeance
    0.79
     speech
    0.79
    Activations Density 0.259%

    No Known Activations