© 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
    2. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    3. GPT2-Small
    4. Residual Similar CE Local
    5. 6-RES_SCL-AJT
    6. 4642
    Prev
    Next
    INDEX
    Explanations

    locations and topics related to a specific city named Hackney

    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
    48
    +0.11
    0.5%
    582
    +0.10
    0.5%
    526
    +0.09
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.11
    0.18
    48
    +0.10
    0.10
    314
    +0.09
    -0.03
    Negative Logits
     occas
    -0.68
    ãħĭ
    -0.67
    ications
    -0.63
    istically
    -0.60
    oice
    -0.60
     circulation
    -0.60
    istic
    -0.59
     confir
    -0.59
    ised
    -0.59
     rul
    -0.57
    POSITIVE LOGITS
    aday
    1.05
    intosh
    1.02
    Hack
    0.96
    athon
    0.89
    ett
    0.86
    enberg
    0.85
    age
    0.82
    ers
    0.79
    hang
    0.76
    Berry
    0.75
    Activations Density 5.513%

    No Known Activations