© 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. 5036
    Prev
    Next
    INDEX
    Explanations

    words related to the concept of injustice or wrongful actions

    oai_token-act-pair · gpt-3.5-turbo

    terms related to injustice and wrongful 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
    gdala
    -0.90
    á
    -0.79
    acid
    -0.70
    phrine
    -0.68
    aden
    -0.68
    ascus
    -0.67
    ingle
    -0.65
    uddy
    -0.65
    annis
    -0.64
    ingo
    -0.64
    POSITIVE LOGITS
    ifiable
    1.43
    ified
    1.31
    icious
    0.92
    ly
    0.85
     unfocusedRange
    0.84
    fully
    0.84
    ãĥ¼ãĥ«
    0.83
     ACTIONS
    0.82
     retribution
    0.79
    eous
    0.78
    Activations Density 0.057%

    No Known Activations