© 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. 35223
    Prev
    Next
    INDEX
    Explanations

    words related to potential abuse or inappropriate behavior, especially concerning grooming and assault

    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
    213
    +0.03
    0.6%
    723
    +0.02
    0.5%
    325
    +0.02
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.03
    0.20
    288
    +0.02
    0.18
    110
    +0.02
    0.08
    Negative Logits
    obin
    -0.17
    aez
    -0.15
    iop
    -0.14
    hell
    -0.13
    ought
    -0.13
    lement
    -0.13
    lance
    -0.13
    ially
    -0.13
    rael
    -0.13
    ector
    -0.13
    POSITIVE LOGITS
     grooming
    0.23
    eering
    0.17
    Sexual
    0.16
    ãħĭãħĭ
    0.15
    ocene
    0.15
     groom
    0.15
    stakes
    0.15
    reditary
    0.15
     Cosponsors
    0.14
    apons
    0.14
    Activations Density 6.672%

    No Known Activations