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. 43299
    Prev
    Next
    INDEX
    Explanations

    possessive forms of nouns

    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
    488
    +0.12
    0.5%
    481
    +0.11
    0.5%
    28
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    28
    +0.12
    0.04
    550
    +0.11
    0.04
    664
    +0.10
    0.04
    Negative Logits
    Reference
    -0.82
    reference
    -0.76
    linked
    -0.71
    tch
    -0.70
    CLAIM
    -0.69
    obin
    -0.65
    lyak
    -0.65
     Engel
    -0.65
    PLIED
    -0.65
    xon
    -0.65
    POSITIVE LOGITS
     skyline
    1.25
     inhabitants
    1.24
     suburbs
    1.18
     residents
    1.18
     streets
    1.13
     neighborhoods
    1.04
     finest
    1.01
     outskirts
    0.98
     mayor
    0.98
     redevelopment
    0.97
    Activations Density 0.136%

    No Known Activations