INDEX

Explanations

superlatives or terms indicating high ranking or importance

oai_token-act-pair · gpt-3.5-turbo

phrases that signify a ranking or categorization of entities as the "greatest" or "best" of all time

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

lav

-0.73

isoft

-0.68

lash

-0.66

ifax

-0.65

ful

-0.64

 Lauder

-0.62

abwe

-0.61

paio

-0.61

kers

-0.60

ritical

-0.60

POSITIVE LOGITS

ocating

0.97

 Invention

0.79

 kinds

0.78

 faiths

0.74

 sorts

0.74

ometry

0.73

ergic

0.72

iance

0.71

ogene

0.70

acea

0.68

Activations Density 0.032%

superlatives or terms indicating high ranking or importance

phrases that signify a ranking or categorization of entities as the "greatest" or "best" of all time

No Comments

No Known Activations