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. 23446
    Prev
    Next
    INDEX
    Explanations

    references to audits and auditing processes

    oai_token-act-pair · gpt-4o-miniTriggered 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
    288
    +0.16
    0.7%
    497
    +0.13
    0.6%
    419
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.16
    0.02
    575
    +0.13
    0.01
    110
    +0.10
    0.01
    Negative Logits
    joy
    -0.82
    ulkan
    -0.82
    axe
    -0.72
    ————————————————
    -0.69
    çīĪ
    -0.67
    Magikarp
    -0.67
     ancest
    -0.66
    spawn
    -0.63
    osi
    -0.62
    hots
    -0.62
    POSITIVE LOGITS
     audit
    1.18
     audits
    0.95
     auditor
    0.85
     reinvest
    0.84
     Audit
    0.82
    onomy
    0.78
     aud
    0.76
     accountant
    0.76
     measure
    0.75
    stration
    0.75
    Activations Density 0.030%

    No Known Activations