INDEX

Explanations

technology-related terms, instructions or specifications

oai_token-act-pair · gpt-3.5-turbo

punctuation and unique formatting elements in the text

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 awaited

-0.68

ascus

-0.64

 seated

-0.63

ãĤ¬

-0.62

arbon

-0.62

intendent

-0.61

 sanct

-0.61

 strap

-0.59

 stamped

-0.58

 surrounded

-0.58

POSITIVE LOGITS

 Including

0.98

 notably

0.86

Examples

0.86

example

0.85

 Examples

0.84

often

0.83

 Sometimes

0.81

usually

0.81

Consider

0.81

 Particularly

0.80

Activations Density 1.087%

technology-related terms, instructions or specifications

punctuation and unique formatting elements in the text

No Comments

No Known Activations