© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    5. GPT2-Small
    7. Residual Similar CE Local
    9. 6-RES_SCL-AJT
    11. 40355
    Prev
    Next
    INDEX
    Explanations

    phrases related to comparisons in the context of technological advancements and market analysis

    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
    447
    +0.22
    1.0%
    197
    +0.13
    0.6%
    641
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.22
    0.31
    393
    +0.13
    0.42
    608
    +0.12
    0.36
    Negative Logits
    Redditor
    -0.81
    Sov
    -0.80
    ships
    -0.69
    Sad
    -0.65
     istg
    -0.65
    rities
    -0.63
    suits
    -0.63
    riot
    -0.62
    mails
    -0.62
    nil
    -0.61
    POSITIVE LOGITS
     regards
    1.31
     regard
    1.14
     terms
    0.95
     respect
    0.89
     mobility
    0.84
     matters
    0.82
     accessibility
    0.77
     Flavoring
    0.75
     direction
    0.72
     quantitative
    0.72
    Activations Density 21.181%

    No Known Activations