© 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. 34016
    Prev
    Next
    INDEX
    Explanations

    alphanumeric patterns within the text

    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
    745
    +0.10
    0.7%
    355
    +0.07
    0.5%
    684
    +0.06
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.10
    0.26
    739
    +0.07
    0.04
    393
    +0.06
    0.27
    Negative Logits
    thood
    -0.53
    ãĥ¼ãĥĨ
    -0.53
    ocene
    -0.50
    netflix
    -0.50
    rongh
    -0.45
    ogyn
    -0.43
    amily
    -0.43
     redes
    -0.43
     Cosponsors
    -0.43
     Orche
    -0.42
    POSITIVE LOGITS
     grabs
    0.46
     abort
    0.44
     elapsed
    0.42
     throws
    0.42
     whence
    0.41
     splits
    0.39
     Snake
    0.37
     yielding
    0.36
     indign
    0.36
     lim
    0.36
    Activations Density 12.383%

    No Known Activations