© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    5. GPT2-Small
    7. Residual Similar CE Local
    9. 6-RES_SCL-AJT
    11. 34816
    Prev
    Next
    INDEX
    Explanations

    discussion related to societal issues such as racism and diverse points of view

    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
    +0.23
    1.1%
    267
    +0.14
    0.7%
    288
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.23
    0.18
    392
    +0.14
    0.09
    588
    +0.12
    0.07
    Negative Logits
    Delivery
    -0.79
    icular
    -0.78
    earchers
    -0.78
    pad
    -0.77
    ership
    -0.75
    ishable
    -0.73
    avez
    -0.73
    TAIN
    -0.72
     largeDownload
    -0.72
    irs
    -0.70
    POSITIVE LOGITS
     slurs
    0.98
     prejudice
    0.93
     racist
    0.87
     racists
    0.82
     sexist
    0.81
     prejud
    0.80
     homophobic
    0.79
     stereotyp
    0.79
     tir
    0.77
     backlash
    0.76
    Activations Density 5.763%

    No Known Activations