© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 3395
    Prev
    Next
    INDEX
    Explanations

    phrases related to personalized recommendations or instructions addressing the reader

    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
    +0.21
    0.7%
    253
    +0.18
    0.6%
    526
    +0.16
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    49
    +0.21
    0.09
    354
    +0.18
    0.08
    270
    +0.16
    0.07
    Negative Logits
    ftime
    -1.10
     Sabha
    -1.06
     Monstrous
    -1.03
    ĸļ
    -1.02
    ĨĴ
    -1.02
    ipal
    -1.01
     Course
    -1.00
     Canaver
    -0.97
    oche
    -0.92
     ammon
    -0.89
    POSITIVE LOGITS
     guys
    1.86
    tub
    1.75
    're
    1.65
    RS
    1.49
    'll
    1.20
     know
    1.17
    NG
    1.15
    Tube
    1.14
    've
    1.13
    lees
    1.12
    Activations Density 1.590%

    No Known Activations