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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 34298
    Prev
    Next
    INDEX
    Explanations

    phrases related to controversies or sensitive topics discussed in public discourse

    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
    +0.18
    0.8%
    725
    +0.12
    0.6%
    323
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    725
    +0.18
    0.10
    199
    +0.12
    0.16
    720
    +0.10
    0.10
    Negative Logits
     Brill
    -0.76
    enment
    -0.70
    ETF
    -0.68
     Agric
    -0.67
    arten
    -0.67
    hess
    -0.67
    gio
    -0.66
     pend
    -0.64
    onder
    -0.64
    td
    -0.63
    POSITIVE LOGITS
     insulting
    1.03
     lewd
    1.00
     slurs
    0.99
     derogatory
    0.98
     insults
    0.96
     dispar
    0.93
     misogyn
    0.91
     joking
    0.90
     harassing
    0.86
     jokes
    0.86
    Activations Density 15.740%

    No Known Activations