INDEX

Explanations

phrases related to deep meanings or interpretations

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "meaning" in various contexts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aliation

-0.75

emetery

-0.67

asting

-0.66

thumbnails

-0.66

Monitor

-0.63

 Breed

-0.63

osterone

-0.62

culosis

-0.62

iets

-0.61

Tur

-0.60

POSITIVE LOGITS

fully

1.61

ful

1.27

fulness

1.26

lessness

1.25

lessly

1.02

FUL

0.94

sworth

0.90

full

0.88

bringer

0.85

istically

0.85

Activations Density 0.028%

phrases related to deep meanings or interpretations

references to the concept of "meaning" in various contexts

No Comments

No Known Activations