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. 4225
    Prev
    Next
    INDEX
    Explanations

    social and political commentary with a critical perspective

    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
    314
    +0.11
    0.6%
    481
    +0.09
    0.5%
    281
    +0.09
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    80
    +0.11
    0.07
    283
    +0.09
    -0.01
    429
    +0.09
    0.05
    Negative Logits
     basin
    -0.60
     specificity
    -0.59
     flats
    -0.54
    sterdam
    -0.53
    Async
    -0.52
     drainage
    -0.50
    akings
    -0.49
    minster
    -0.49
    PN
    -0.49
    Ru
    -0.49
    POSITIVE LOGITS
    rencies
    0.67
    insk
    0.66
    ãĥīãĥ©
    0.65
    ÃŁ
    0.65
    ifiable
    0.64
    thinkable
    0.62
    sbm
    0.62
    apon
    0.61
    isable
    0.61
    士
    0.59
    Activations Density 13.513%

    No Known Activations