© 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. 10-RES-JB
    11. 6897
    Prev
    Next
    INDEX
    Explanations

    adverbs describing the intensity of an attribute or action

    oai_token-act-pair · gpt-3.5-turbo

    expressions of degree or intensity in descriptions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    lessness
    -0.78
    LET
    -0.78
    moil
    -0.76
    raltar
    -0.75
    only
    -0.75
     Cannot
    -0.71
    rosis
    -0.71
    çļ
    -0.69
    tera
    -0.67
    kamp
    -0.66
    POSITIVE LOGITS
     conducive
    1.23
     glamorous
    1.16
     satisfactory
    1.09
     attractive
    1.04
     desirable
    1.04
     pleasant
    1.04
     useful
    1.00
     reliable
    1.00
     productive
    1.00
     appealing
    1.00
    Activations Density 0.074%

    No Known Activations