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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 45193
    Prev
    Next
    INDEX
    Explanations

    specific words related to architectural and geographical locations

    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
    232
    +0.12
    0.5%
    314
    +0.12
    0.5%
    739
    +0.11
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.12
    0.21
    232
    +0.12
    0.10
    314
    +0.11
    -0.04
    Negative Logits
    anyahu
    -0.73
    ppo
    -0.70
     Romeo
    -0.70
    ibaba
    -0.70
     flora
    -0.68
    obar
    -0.67
     behold
    -0.66
    ocal
    -0.65
     fertility
    -0.65
     cartel
    -0.65
    POSITIVE LOGITS
    ozy
    1.26
    ansas
    1.15
    enment
    1.10
    eting
    1.01
    ness
    0.99
    ling
    0.97
    itect
    0.97
    otle
    0.96
    ening
    0.96
    anian
    0.95
    Activations Density 7.515%

    No Known Activations