© 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. 4110
    Prev
    Next
    INDEX
    Explanations

    descriptions related to circular structures or regions

    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
    +0.21
    0.8%
    326
    +0.20
    0.8%
    647
    +0.18
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.21
    0.17
    647
    +0.20
    0.06
    381
    +0.18
    0.07
    Negative Logits
    ufact
    -1.01
    essors
    -0.87
    icago
    -0.86
    UGE
    -0.85
    irrel
    -0.83
    éĹĺ
    -0.83
    tical
    -0.81
    EMP
    -0.81
     Fiorina
    -0.79
     autop
    -0.78
    POSITIVE LOGITS
    tone
    1.62
    leader
    1.46
    tones
    1.43
    leaders
    1.41
    worm
    1.24
    git
    1.23
    wra
    1.19
    rings
    1.11
    mail
    1.05
    bard
    1.02
    Activations Density 6.157%

    No Known Activations