© 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. 12880
    Prev
    Next
    INDEX
    Explanations

    mentions of the word "ghost" in various contexts

    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
    +0.17
    0.8%
    571
    +0.14
    0.6%
    326
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.17
    0.08
    476
    +0.14
    0.03
    270
    +0.13
    0.03
    Negative Logits
    unes
    -0.75
    ×IJ
    -0.73
    enegger
    -0.73
    ibaba
    -0.72
    thritis
    -0.72
    utics
    -0.72
    NRS
    -0.72
    tics
    -0.71
    erity
    -0.71
    onian
    -0.71
    POSITIVE LOGITS
    busters
    1.14
    writer
    0.98
    buster
    0.95
    written
    0.88
    door
    0.86
    writing
    0.86
     haunting
    0.85
    ly
    0.82
    writers
    0.79
     Recon
    0.77
    Activations Density 8.050%

    No Known Activations