© 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. 8-RES-JB
    11. 6507
    Prev
    Next
    INDEX
    Explanations

    terms or phrases that are persuasive or convincing in nature

    oai_token-act-pair · gpt-3.5-turbo

    the concept of compelling arguments or reasons, particularly in relation to urgency or important issues

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    antam
    -0.72
    alde
    -0.71
    strap
    -0.68
    adders
    -0.66
     Columb
    -0.65
    aunder
    -0.64
    hops
    -0.64
    lopp
    -0.62
    aido
    -0.61
    eties
    -0.61
    POSITIVE LOGITS
     urgent
    0.84
     urgency
    0.83
     imperative
    0.79
     proble
    0.79
    need
    0.77
    LY
    0.75
     compel
    0.75
     demand
    0.73
     necessity
    0.72
     urge
    0.72
    Activations Density 0.091%

    No Known Activations