© 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. 17327
    Prev
    Next
    INDEX
    Explanations

    terms related to persuasion or convincing someone

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to persuasion and convincing others

    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
    emale
    -0.73
    practice
    -0.73
    vertisement
    -0.72
    illary
    -0.70
     largeDownload
    -0.69
    eatures
    -0.68
    bard
    -0.68
    nam
    -0.66
    served
    -0.66
    apped
    -0.65
    POSITIVE LOGITS
     skeptics
    0.93
     policymakers
    0.81
     otherwise
    0.80
    ingly
    0.79
     others
    0.79
     lawmakers
    0.78
     voters
    0.78
     enough
    0.78
     reluctant
    0.75
     congress
    0.74
    Activations Density 0.030%

    No Known Activations