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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 39745
    Prev
    Next
    INDEX
    Explanations

    statements expressing delight, pleasure, or patriotism

    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
    +0.14
    0.7%
    665
    +0.11
    0.5%
    746
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    49
    +0.14
    0.19
    608
    +0.11
    0.22
    718
    +0.11
    0.20
    Negative Logits
    soDeliveryDate
    -0.77
     millenn
    -0.70
     wills
    -0.65
     objectionable
    -0.62
    hur
    -0.62
     Advertisement
    -0.57
    limits
    -0.56
     tant
    -0.56
    calling
    -0.56
     overriding
    -0.55
    POSITIVE LOGITS
     announce
    1.01
     learn
    0.98
     congratulate
    0.90
     admit
    0.88
     celebrate
    0.88
     receive
    0.85
     remind
    0.85
     find
    0.84
     hear
    0.83
     see
    0.83
    Activations Density 7.145%

    No Known Activations