© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 12288
    9. 8-RES_FS12288-JB
    11. 1440
    Prev
    Next
    INDEX
    Explanations

    news articles and specific web links

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    107
    +0.13
    0.6%
    331
    +0.11
    0.5%
    587
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    107
    +0.13
    0.04
    409
    +0.11
    0.04
    586
    +0.10
    0.03
    Negative Logits
     compounded
    -0.68
     Parenthood
    -0.62
     Belt
    -0.61
     Athletic
    -0.61
     Bradley
    -0.60
     IST
    -0.60
     Oval
    -0.59
     Daly
    -0.59
     Dull
    -0.58
     Rath
    -0.58
    POSITIVE LOGITS
    /,
    1.32
    /)
    1.25
    /?
    1.21
    /.
    1.18
    /#
    1.10
    /_
    1.08
    _
    1.06
    php
    1.05
    detail
    1.01
    /
    0.98
    Activations Density 0.141%

    No Known Activations