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. 25562
    Prev
    Next
    INDEX
    Explanations

    phrases that emphasize the quantity or significance of an entity, particularly in terms of totals or counts

    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
    393
    +0.12
    0.6%
    362
    +0.10
    0.5%
    289
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    223
    +0.12
    0.01
    289
    +0.10
    0.01
    748
    +0.09
    0.01
    Negative Logits
    pring
    -0.63
     Presence
    -0.61
     condem
    -0.60
    nesses
    -0.56
    era
    -0.56
     trave
    -0.55
    tera
    -0.54
    pod
    -0.54
    ness
    -0.53
    adays
    -0.53
    POSITIVE LOGITS
     eight
    1.05
     seven
    1.00
     six
    1.00
     nine
    0.97
     thirteen
    0.97
     THREE
    0.93
     fourteen
    0.93
     nineteen
    0.91
     four
    0.89
     seventy
    0.89
    Activations Density 0.020%

    No Known Activations