© 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. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 4-RES-JB
    11. 10536
    Prev
    Next
    INDEX
    Explanations

    terms related to drinking or substances to consume

    oai_token-act-pair · gpt-3.5-turbo

    references to substance use, particularly with a focus on 'drugs'

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 4-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.4.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    4
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ERN
    -0.85
    eers
    -0.80
     Tactics
    -0.75
     Pose
    -0.72
     Corpus
    -0.70
    eering
    -0.68
     Territories
    -0.65
     Leilan
    -0.65
    orkshire
    -0.64
     Blueprint
    -0.64
    POSITIVE LOGITS
    inking
    1.23
    unk
    1.14
    inker
    1.14
    inks
    1.12
    agons
    1.11
    agnar
    1.10
    ink
    1.01
    ifted
    0.95
    ifts
    0.94
    ained
    0.93
    Activations Density 0.012%

    No Known Activations