© 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. 45741
    Prev
    Next
    INDEX
    Explanations

    language related to technology and commerce

    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
    +0.22
    1.1%
    393
    +0.11
    0.6%
    583
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    583
    +0.22
    0.06
    382
    +0.11
    0.07
    86
    +0.11
    0.06
    Negative Logits
    atform
    -0.58
    oud
    -0.57
     Cheong
    -0.56
     Discrimination
    -0.56
    maker
    -0.56
    ļéĨĴ
    -0.56
    ob
    -0.55
    enegger
    -0.55
    jandro
    -0.54
    forth
    -0.54
    POSITIVE LOGITS
     sake
    0.84
     curious
    0.82
     wishing
    0.81
     redes
    0.80
     unfamiliar
    0.79
     interested
    0.78
     purposes
    0.78
     wanting
    0.76
     adventurous
    0.74
     unlucky
    0.71
    Activations Density 7.335%

    No Known Activations