© 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. 11103
    Prev
    Next
    INDEX
    Explanations

    phrases related to expressing strong emotion or urgency

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of quoted speech

    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
     hardened
    -0.83
     affected
    -0.80
     targeted
    -0.80
     hosts
    -0.80
     frontline
    -0.80
     isolated
    -0.79
     distinguished
    -0.78
     backlog
    -0.76
     foreground
    -0.76
     vigil
    -0.76
    POSITIVE LOGITS
    Hey
    1.78
    Fuck
    1.73
    Oh
    1.67
    Yeah
    1.66
    hello
    1.62
    Hello
    1.60
    yeah
    1.60
    fuck
    1.60
    sorry
    1.59
    please
    1.59
    Activations Density 0.112%

    No Known Activations