INDEX

Explanations

mentions of coffee-related terms

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "coffee" and related expressions

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

oppable

-0.80

ROR

-0.79

orse

-0.74

 Caldwell

-0.72

 newcom

-0.68

yss

-0.68

ollow

-0.67

Filename

-0.66

veh

-0.65

Population

-0.65

POSITIVE LOGITS

 beans

1.20

 Beans

1.10

 bean

1.01

 brewed

1.01

 drip

0.96

 coffee

0.94

weed

0.90

 cups

0.90

cake

0.90

 drinkers

0.88

Activations Density 0.015%

mentions of coffee-related terms

occurrences of the word "coffee" and related expressions

No Comments

No Known Activations