© 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 E2E Recon
    9. 6-RES_SCEFR-AJT
    11. 1348
    Prev
    Next
    INDEX
    Explanations

    phrases related to specific skills or abilities, such as how to do something

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    neuronpedia/gpt2-small__res_scefr-ajt/6-res_scefr-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
    393
    +0.23
    0.7%
    644
    +0.14
    0.4%
    55
    +0.13
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.23
    0.06
    242
    +0.14
    0.04
    750
    +0.13
    0.04
    Negative Logits
    etheus
    -1.15
    lished
    -1.04
    eworthy
    -1.02
    aredevil
    -1.00
    rarily
    -0.98
     sugg
    -0.98
     ect
    -0.98
    glomer
    -0.95
    byss
    -0.91
    iversal
    -0.91
    POSITIVE LOGITS
    RECT
    0.98
    Tact
    0.91
     drill
    0.90
    assy
    0.90
    %-
    0.87
     stroke
    0.87
     kind
    0.85
     firsthand
    0.85
     lick
    0.85
     weakness
    0.85
    Activations Density 0.495%

    No Known Activations