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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 45769
    Prev
    Next
    INDEX
    Explanations

    sensitive topics and mentions of violence

    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
    373
    +0.27
    1.3%
    640
    +0.13
    0.6%
    228
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.27
    0.09
    94
    +0.13
    0.07
    373
    +0.10
    -0.01
    Negative Logits
    heit
    -0.87
    cius
    -0.75
    ãĤ¼ãĤ¦ãĤ¹
    -0.70
    boro
    -0.64
     Stellar
    -0.60
    icht
    -0.57
    utical
    -0.57
     CoC
    -0.56
     bitcoins
    -0.56
    minist
    -0.56
    POSITIVE LOGITS
    Els
    0.80
    Known
    0.65
    Numbers
    0.65
    Such
    0.64
    Bus
    0.64
     Jamal
    0.64
    adder
    0.63
    Labour
    0.63
    However
    0.63
    Jam
    0.63
    Activations Density 0.197%

    No Known Activations