© 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. 8951
    Prev
    Next
    INDEX
    Explanations

    reasons or explanations for certain actions or phenomena

    oai_token-act-pair · gpt-3.5-turbo

    inquiries about reasons or explanations for feelings and actions

    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
    lator
    -0.87
    ymph
    -0.79
    aughed
    -0.75
    iece
    -0.73
     Roller
    -0.70
    wana
    -0.65
    ibaba
    -0.65
    robe
    -0.61
    iHUD
    -0.61
    opic
    -0.60
    POSITIVE LOGITS
    soever
    0.95
     exactly
    0.78
     why
    0.74
     they
    0.74
     bother
    0.72
    abl
    0.72
     people
    0.70
     someone
    0.67
     somebody
    0.67
     WHY
    0.64
    Activations Density 0.043%

    No Known Activations