© 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. 4168
    Prev
    Next
    INDEX
    Explanations

    phrases related to decisions not to do something

    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
    +0.24
    1.1%
    651
    +0.10
    0.5%
    665
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    307
    +0.24
    0.12
    221
    +0.10
    0.08
    360
    +0.09
    0.11
    Negative Logits
    ixel
    -0.65
     Springs
    -0.60
    ãĤ¼ãĤ¦ãĤ¹
    -0.59
    itor
    -0.57
     Fountain
    -0.55
    CSS
    -0.55
     Expansion
    -0.54
    izational
    -0.53
     Journalism
    -0.53
     Appears
    -0.53
    POSITIVE LOGITS
    icably
    1.14
    epad
    0.92
     necessarily
    0.91
    withstanding
    0.87
    hin
    0.81
    icable
    0.78
    externalActionCode
    0.72
     surpr
    0.70
    ifying
    0.68
     criminally
    0.67
    Activations Density 6.245%

    No Known Activations