© 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. 33929
    Prev
    Next
    INDEX
    Explanations

    references to images with Getty credits

    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
    409
    +0.10
    0.5%
    741
    +0.10
    0.5%
    665
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    385
    +0.10
    0.16
    640
    +0.10
    0.22
    308
    +0.10
    0.06
    Negative Logits
    rone
    -0.65
     taxpayers
    -0.63
     contrace
    -0.63
     "$:/
    -0.62
     Loll
    -0.60
    obos
    -0.59
    lake
    -0.59
    steen
    -0.59
     outweigh
    -0.59
     guiActiveUn
    -0.59
    POSITIVE LOGITS
    fecture
    0.84
    Getty
    0.78
    rency
    0.73
    senal
    0.68
    ĸļ
    0.66
    unal
    0.65
    Balt
    0.64
    Sah
    0.63
    rahim
    0.61
    execute
    0.61
    Activations Density 10.037%

    No Known Activations