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. 41268
    Prev
    Next
    INDEX
    Explanations

    street names and related address components

    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
    256
    +0.12
    0.5%
    103
    +0.11
    0.5%
    59
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.12
    0.04
    220
    +0.11
    0.03
    99
    +0.10
    0.02
    Negative Logits
    Analy
    -0.67
    agging
    -0.65
     causation
    -0.63
     thirst
    -0.63
    umbn
    -0.63
    auc
    -0.62
    Unknown
    -0.61
     freeze
    -0.61
    ctica
    -0.59
    anything
    -0.59
    POSITIVE LOGITS
     Blvd
    1.26
     Ave
    1.07
     Rd
    1.03
     acres
    0.84
     Lane
    0.77
     RD
    0.72
    illion
    0.72
     Oaks
    0.70
     Place
    0.70
     Drive
    0.70
    Activations Density 0.186%

    No Known Activations