© 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. 33962
    Prev
    Next
    INDEX
    Explanations

    locations within a scene

    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
    212
    +0.11
    0.5%
    26
    +0.11
    0.4%
    571
    +0.11
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    571
    +0.11
    0.07
    422
    +0.11
    0.04
    212
    +0.11
    0.05
    Negative Logits
    entric
    -0.86
    aukee
    -0.86
    inational
    -0.84
    eteria
    -0.84
    ifix
    -0.82
    ensable
    -0.79
    izo
    -0.79
    upon
    -0.78
    onential
    -0.78
    ourning
    -0.78
    POSITIVE LOGITS
    side
    1.53
    packs
    1.30
    woods
    1.28
    GROUND
    1.26
    seat
    1.21
    door
    1.17
    country
    1.15
    stab
    1.13
    stories
    1.11
    hand
    1.11
    Activations Density 4.026%

    No Known Activations