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. 34452
    Prev
    Next
    INDEX
    Explanations

    mentions of geopolitical events and politics

    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
    +0.13
    0.5%
    248
    +0.12
    0.5%
    200
    +0.11
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    752
    +0.13
    0.17
    57
    +0.12
    0.10
    659
    +0.11
    0.13
    Negative Logits
    inas
    -0.78
     Title
    -0.76
    atown
    -0.76
    antry
    -0.75
     VID
    -0.73
    achine
    -0.71
     Shall
    -0.70
     basics
    -0.69
    govtrack
    -0.68
    Owner
    -0.67
    POSITIVE LOGITS
     Embassy
    1.02
     embassy
    1.01
     diplomats
    0.99
     diplomatic
    0.91
     imperialism
    0.90
    AAF
    0.87
     missionaries
    0.85
     neocons
    0.84
     consulate
    0.84
    jen
    0.82
    Activations Density 13.870%

    No Known Activations