INDEX

Explanations

plural nouns that refer to various things or concepts

oai_token-act-pair · gpt-3.5-turbo

phrases indicating groups or categories of things

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

racuse

-0.78

osate

-0.77

atown

-0.76

slaught

-0.75

rium

-0.71

 Shutterstock

-0.70

opus

-0.69

abus

-0.69

ipl

-0.67

edia

-0.67

POSITIVE LOGITS

 factors

1.22

 extremes

1.16

 phenomena

1.12

 scenarios

1.02

 situations

0.99

 propositions

0.99

 endeavors

0.96

 categories

0.96

 modes

0.94

 genres

0.93

Activations Density 0.135%

plural nouns that refer to various things or concepts

phrases indicating groups or categories of things

No Comments

No Known Activations