© 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. 114
    Prev
    Next
    INDEX
    Explanations

    names, particularly the name "Ross"

    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
    195
    +0.16
    0.6%
    682
    +0.14
    0.5%
    25
    +0.13
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.16
    0.12
    288
    +0.14
    0.13
    682
    +0.13
    0.08
    Negative Logits
    glers
    -1.05
    brance
    -0.98
    urated
    -0.94
    CHO
    -0.93
    ILE
    -0.93
    ificant
    -0.91
    mination
    -0.91
    lihood
    -0.87
     [|
    -0.87
    service
    -0.87
    POSITIVE LOGITS
    oulos
    1.11
    anches
    1.00
    bach
    0.98
    aunders
    0.98
    etti
    0.93
    ushima
    0.91
     Ross
    0.90
     Shepherd
    0.88
    ionics
    0.88
     Ul
    0.87
    Activations Density 1.965%

    No Known Activations