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. 37195
    Prev
    Next
    INDEX
    Explanations

    calls to action or prompts for audience engagement

    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
    566
    +0.12
    0.6%
    326
    +0.10
    0.5%
    530
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.12
    0.02
    271
    +0.10
    0.02
    176
    +0.10
    0.02
    Negative Logits
    night
    -0.75
    stood
    -0.73
    ised
    -0.71
    pires
    -0.69
    ties
    -0.69
    nia
    -0.68
    fired
    -0.66
    rays
    -0.65
    ises
    -0.65
    lets
    -0.63
    POSITIVE LOGITS
    MORE
    1.39
    BOOK
    1.38
    ING
    1.37
    HEAD
    1.27
     ALSO
    1.26
    ERSON
    1.25
    MAN
    1.24
    PLAY
    1.23
    DOWN
    1.23
    WATCH
    1.23
    Activations Density 0.030%

    No Known Activations