© 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. 14019
    Prev
    Next
    INDEX
    Explanations

    phrases related to secretive or unethical activities

    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
    +0.19
    0.9%
    288
    +0.11
    0.5%
    326
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    85
    +0.19
    0.10
    326
    +0.11
    0.20
    526
    +0.11
    -0.06
    Negative Logits
    */(
    -0.88
    yip
    -0.82
    HCR
    -0.80
    ajor
    -0.76
    iphate
    -0.76
    ommel
    -0.74
    ISC
    -0.74
    uther
    -0.74
    iculture
    -0.73
    icultural
    -0.72
    POSITIVE LOGITS
     laundry
    0.99
     linen
    0.84
     dirty
    0.81
     tricks
    0.79
     diaper
    0.76
    bag
    0.75
     mole
    0.74
     diapers
    0.74
     stains
    0.67
    dirty
    0.66
    Activations Density 7.514%

    No Known Activations