Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    3. GPT2-Small
    4. Residual Similar CE Local
    5. 6-RES_SCL-AJT
    6. 4594
    Prev
    Next
    INDEX
    Explanations

    descriptions of military or geopolitical situations

    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
    +0.25
    1.1%
    393
    +0.18
    0.8%
    274
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.25
    0.22
    598
    +0.18
    0.09
    563
    +0.12
    0.14
    Negative Logits
    "],"
    -0.78
    natureconservancy
    -0.75
    thro
    -0.74
    emis
    -0.69
    hips
    -0.68
    ãĥĥãĤ¯
    -0.66
    minent
    -0.65
    uty
    -0.65
    ogly
    -0.65
    atur
    -0.64
    POSITIVE LOGITS
    cher
    0.94
     mattered
    0.87
     culminated
    0.84
     pesky
    0.82
     translates
    0.81
     sounds
    0.81
     includes
    0.81
     same
    0.80
     begs
    0.80
     resulted
    0.79
    Activations Density 9.657%

    No Known Activations