© 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. 34907
    Prev
    Next
    INDEX
    Explanations

    phrases related to announcements or statements of necessity

    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%
    285
    +0.15
    0.6%
    651
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    285
    +0.22
    0.05
    635
    +0.15
    0.06
    211
    +0.11
    -0.01
    Negative Logits
     Surv
    -0.81
     rabbit
    -0.77
    imore
    -0.75
     Sabb
    -0.71
     Alb
    -0.70
     corrid
    -0.67
     Dice
    -0.67
    é¾
    -0.66
     precincts
    -0.65
     ank
    -0.65
    POSITIVE LOGITS
    lessly
    1.48
    lest
    0.92
    iness
    0.87
    kward
    0.84
    ful
    0.82
    onies
    0.81
    lessness
    0.79
    iac
    0.78
    iest
    0.77
    ulkan
    0.77
    Activations Density 7.396%

    No Known Activations