INDEX

Explanations

phrases indicating movement or action

oai_token-act-pair · gpt-3.5-turbo

instances of the word "go."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ificent

-0.74

ament

-0.74

ullah

-0.74

uctor

-0.71

itionally

-0.71

 Horus

-0.70

creen

-0.68

icio

-0.64

oret

-0.61

orsi

-0.60

POSITIVE LOGITS

vt

1.13

lems

1.04

verning

0.94

ggle

0.90

ogly

0.83

aded

0.83

ALK

0.81

etz

0.81

Ń·

0.80

 overboard

0.80

Activations Density 0.106%

phrases indicating movement or action

instances of the word "go."

No Comments

No Known Activations