© 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. 40560
    Prev
    Next
    INDEX
    Explanations

    unexpected thoughts or scenarios

    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
    +0.14
    0.7%
    460
    +0.08
    0.4%
    287
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.14
    0.22
    393
    +0.08
    0.30
    307
    +0.08
    0.09
    Negative Logits
    adra
    -0.68
    nee
    -0.65
    berry
    -0.60
    utic
    -0.56
    gdala
    -0.55
    ioxide
    -0.54
    igl
    -0.53
    adj
    -0.53
    Dialogue
    -0.53
    &
    -0.53
    POSITIVE LOGITS
     nor
    0.77
     anymore
    0.67
     slightest
    0.64
    Initialized
    0.63
     bothered
    0.62
     anything
    0.62
     existed
    0.60
    DERR
    0.59
     EVER
    0.58
    existent
    0.58
    Activations Density 15.854%

    No Known Activations