© 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 Local
    9. 6-RES_SCL-AJT
    11. 40825
    Prev
    Next
    INDEX
    Explanations

    actions or decisions made in a specific situation or context

    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
    +0.26
    1.6%
    393
    +0.09
    0.6%
    400
    +0.08
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.26
    0.33
    14
    +0.09
    0.10
    473
    +0.08
    0.08
    Negative Logits
    bard
    -0.58
    PI
    -0.57
    dar
    -0.55
    oS
    -0.54
    kat
    -0.51
    mma
    -0.51
    undo
    -0.50
    kay
    -0.49
    Paris
    -0.49
    arez
    -0.49
    POSITIVE LOGITS
     nonetheless
    0.90
    etheless
    0.82
     nevertheless
    0.79
     persists
    0.64
     persisted
    0.64
     beware
    0.60
     alas
    0.59
     sheer
    0.55
     specifics
    0.54
    theless
    0.51
    Activations Density 15.631%

    No Known Activations