© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 35405
    Prev
    Next
    INDEX
    Explanations

    phrases expressing personal opinions or feelings, specifically emphasizing words like "honestly" and "frankly"

    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
    +0.22
    0.9%
    88
    +0.13
    0.5%
    99
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    377
    +0.22
    0.21
    314
    +0.13
    -0.05
    6
    +0.12
    0.17
    Negative Logits
     Landing
    -0.85
    indal
    -0.77
    aeus
    -0.76
    anwhile
    -0.73
     mobilization
    -0.73
    tein
    -0.72
     activation
    -0.71
    cair
    -0.71
     curfew
    -0.71
    ilater
    -0.71
    POSITIVE LOGITS
     speaking
    0.92
    âĸijâĸij
    0.84
    zers
    0.82
    é¾įåĸļ士
    0.81
    NESS
    0.76
    UGE
    0.76
    ×Ļ
    0.74
     honestly
    0.73
    âĸij
    0.73
    BuyableInstoreAndOnline
    0.72
    Activations Density 10.654%

    No Known Activations