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 49152
    5. 8-RES_FS49152-JB
    6. 26351
    Prev
    Next
    INDEX
    Explanations

    forms of the word "fail" or related words indicating failure or disappointment

    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
    +0.14
    0.6%
    436
    +0.12
    0.5%
    251
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    436
    +0.14
    0.01
    251
    +0.12
    0.01
    746
    +0.11
    0.01
    Negative Logits
    ãĥ¯ãĥ³
    -0.63
    wcs
    -0.62
     reper
    -0.61
     distingu
    -0.59
    #$
    -0.59
    perty
    -0.59
     comprom
    -0.58
    etheless
    -0.58
     fortified
    -0.58
    perature
    -0.57
    POSITIVE LOGITS
    escription
    0.86
    irection
    0.85
     DMCA
    0.81
    ãĤ¶
    0.74
    river
    0.74
    ctions
    0.74
    dit
    0.74
    lag
    0.73
    ategor
    0.72
    gency
    0.70
    Activations Density 0.006%

    No Known Activations