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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 1-RES-JB
    11. 4845
    Prev
    Next
    INDEX
    Explanations

    phrases related to medical prescriptions or instructions

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "dispense" and its variations, indicating a focus on distribution or provision of substances

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Reviewer
    -1.00
    mere
    -0.78
    sen
    -0.78
    66666666
    -0.75
    lihood
    -0.73
    HCR
    -0.73
    wikipedia
    -0.72
    tle
    -0.71
    rium
    -0.70
    FORE
    -0.70
    POSITIVE LOGITS
     disp
    1.15
    ersion
    0.99
    ables
    0.96
    ensed
    0.93
    abled
    0.93
    ossession
    0.88
    ennes
    0.86
    orter
    0.84
    ersive
    0.81
    assion
    0.81
    Activations Density 0.007%

    No Known Activations