INDEX

Explanations

instances of introducing or being introduced to something

oai_token-act-pair · gpt-3.5-turbo

phrases related to the act of introducing concepts, people, or changes

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

unker

-0.81

ache

-0.75

enter

-0.71

halla

-0.68

ogun

-0.67

alyses

-0.67

fall

-0.66

iac

-0.65

opal

-0.64

ocaly

-0.64

POSITIVE LOGITS

into

0.82

 into

0.80

new

0.79

 strains

0.77

 legislation

0.76

 INTO

0.71

 introduce

0.71

 stricter

0.70

 novel

0.68

 compulsory

0.68

Activations Density 0.051%

instances of introducing or being introduced to something

phrases related to the act of introducing concepts, people, or changes

No Comments

No Known Activations