INDEX

Explanations

instances of the word "double"

oai_token-act-pair · gpt-3.5-turbo

references to "double" in various contexts

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

omo

-0.81

ONY

-0.73

ega

-0.68

onite

-0.67

utical

-0.67

lov

-0.66

ihara

-0.65

istry

-0.64

uria

-0.64

laus

-0.61

POSITIVE LOGITS

header

1.14

headers

1.05

 sided

0.96

 digit

0.91

 Decker

0.87

 edged

0.87

 dipping

0.83

 digits

0.83

 jeopardy

0.80

think

0.80

Activations Density 0.031%

instances of the word "double"

references to "double" in various contexts

No Comments

No Known Activations