© 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
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 11-RES-JB
    6. 15700
    Prev
    Next
    INDEX
    Explanations

    words related to personal experiences and emotions

    oai_token-act-pair · gpt-3.5-turbo

    phrases that describe distressing or uncomfortable situations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ranch
    -0.89
    abet
    -0.78
    riber
    -0.76
    Resp
    -0.74
    luaj
    -0.71
    ospons
    -0.70
    orsi
    -0.69
    soever
    -0.69
    raper
    -0.68
    backer
    -0.67
    POSITIVE LOGITS
     unbearable
    1.30
     heartbreaking
    1.22
     unbelievable
    1.21
     humiliating
    1.21
     horrible
    1.21
     disgusting
    1.20
     scary
    1.19
     embarrassing
    1.18
     surreal
    1.17
     terrifying
    1.14
    Activations Density 0.175%

    No Known Activations