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. 716
    Prev
    Next
    INDEX
    Explanations

    proper nouns and book titles

    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
    6
    +0.12
    0.5%
    232
    +0.11
    0.5%
    476
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.12
    0.06
    6
    +0.11
    0.05
    326
    +0.10
    0.04
    Negative Logits
     referen
    -0.92
     satell
    -0.83
    ivably
    -0.77
     millenn
    -0.76
     carbohyd
    -0.75
     rall
    -0.74
     practition
    -0.73
    ahon
    -0.72
     endors
    -0.71
    aspers
    -0.71
    POSITIVE LOGITS
     Revis
    1.18
     Chronicles
    1.18
     Trilogy
    1.16
     Reloaded
    1.15
     Massacre
    1.10
     Remastered
    1.10
    !,
    1.07
     Pt
    1.05
     Redux
    1.02
     Vampire
    1.01
    Activations Density 0.174%

    No Known Activations