INDEX

Explanations

double-quotes followed by unique, non-alphanumeric characters

oai_token-act-pair · gpt-3.5-turbo

dialogues or exchanges in conversational formats

oai_token-act-pair · gpt-4o-mini Triggered 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

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 confir

-0.96

 proport

-0.89

 migration

-0.87

 moder

-0.84

 govern

-0.83

 favor

-0.83

 enriched

-0.79

 divest

-0.79

xual

-0.79

 funnel

-0.79

POSITIVE LOGITS

Suddenly

1.92

"âĢ¦

1.77

"...

1.72

"-

1.67

Chapter

1.66

"'

1.62

Behind

1.61

CHAPTER

1.61

Seeing

1.59

Herm

1.58

Activations Density 0.163%

double-quotes followed by unique, non-alphanumeric characters

dialogues or exchanges in conversational formats

No Comments

No Known Activations