© 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
    2. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    3. GPT2-Small
    4. Residual Similar CE E2E Recon
    5. 6-RES_SCEFR-AJT
    6. 5961
    Prev
    Next
    INDEX
    Explanations

    mathematical symbols and equations

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    neuronpedia/gpt2-small__res_scefr-ajt/6-res_scefr-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
    178
    +0.15
    0.5%
    728
    +0.14
    0.5%
    699
    +0.13
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    208
    +0.15
    0.03
    514
    +0.14
    0.02
    683
    +0.13
    0.02
    Negative Logits
    thin
    -0.89
    inki
    -0.89
    kson
    -0.86
    unin
    -0.84
     horizont
    -0.80
     laund
    -0.80
    AAF
    -0.79
     minim
    -0.79
     cad
    -0.78
     seism
    -0.77
    POSITIVE LOGITS
    Name
    1.08
    display
    1.03
    inline
    0.99
    ĺ
    0.92
     WATCHED
    0.91
    inez
    0.88
    name
    0.85
    note
    0.85
    Error
    0.83
    feature
    0.83
    Activations Density 0.241%

    No Known Activations