© 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. 3877
    Prev
    Next
    INDEX
    Explanations

    phrases related to content warnings or alerts

    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
    +0.27
    0.9%
    566
    +0.20
    0.7%
    55
    +0.20
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    55
    +0.27
    0.17
    102
    +0.20
    0.16
    566
    +0.20
    0.14
    Negative Logits
    isin
    -0.96
     Might
    -0.92
     showc
    -0.90
     Patrol
    -0.85
    strous
    -0.85
     brace
    -0.83
     Bars
    -0.82
    iop
    -0.81
    éĹĺ
    -0.80
    naire
    -0.80
    POSITIVE LOGITS
     able
    1.66
     bothered
    1.53
     anymore
    1.32
     forgiven
    1.30
     anywhere
    1.27
     necessarily
    1.25
     counted
    1.24
     harmed
    1.21
     adequately
    1.19
     entirely
    1.18
    Activations Density 1.830%

    No Known Activations