Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    3. GPT2-Small
    4. Residual Similar CE Local
    5. 6-RES_SCL-AJT
    6. 45347
    Prev
    Next
    INDEX
    Explanations

    mentions of a specific type of rock - granite

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    neuronpedia/gpt2-small__res_scl-ajt/6-res_scl-ajt
    Prompts (Dashboard)
    12,288 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    46,080
    Data Type
    torch.float32
    Hook Point
    blocks.6.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    apollo-research/Skylion007-openwebtext-tokenizer-gpt2
    Hook Point Layer
    6
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    326
    +0.12
    0.6%
    447
    +0.11
    0.6%
    421
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.12
    0.24
    577
    +0.11
    0.12
    683
    +0.11
    0.09
    Negative Logits
     Mutual
    -0.65
    payer
    -0.62
    åº
    -0.60
     Gutenberg
    -0.60
     Lovecraft
    -0.60
     Laugh
    -0.60
     cele
    -0.58
    yers
    -0.57
    suit
    -0.57
    lessly
    -0.56
    POSITIVE LOGITS
    ules
    1.02
    ular
    0.99
    ularity
    0.94
    vik
    0.86
    agraph
    0.75
    ulates
    0.74
    adier
    0.74
    izo
    0.74
    igans
    0.72
    oline
    0.72
    Activations Density 9.181%

    No Known Activations