INDEX

Explanations

phrases related to maximizing or reaching the maximum level of something

oai_token-act-pair · gpt-3.5-turbo

references to numerical maxima or limits

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

FORE

-0.72

CHO

-0.68

 Moons

-0.67

 Shib

-0.67

tty

-0.66

bare

-0.66

 Forth

-0.66

 Patent

-0.66

 Guys

-0.66

RECT

-0.65

POSITIVE LOGITS

imus

1.22

max

0.97

max

0.96

imum

0.93

imize

0.90

icum

0.86

imates

0.85

 veter

0.84

nesota

0.84

eport

0.82

Activations Density 0.005%

phrases related to maximizing or reaching the maximum level of something

references to numerical maxima or limits

No Comments

No Known Activations