INDEX

Explanations

adjectives describing quantities or intensities

oai_token-act-pair · gpt-3.5-turbo

words ending in 'ful'

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

IRO

-0.75

ARC

-0.69

BAT

-0.67

IPP

-0.67

Reviewer

-0.63

ISO

-0.62

mercial

-0.61

NEC

-0.61

Downloadha

-0.61

utherford

-0.61

POSITIVE LOGITS

vous

0.87

terday

0.87

filled

0.86

poons

0.83

roxy

0.79

ths

0.77

eners

0.76

rous

0.74

lio

0.74

tons

0.72

Activations Density 0.015%

adjectives describing quantities or intensities

words ending in 'ful'

No Comments

No Known Activations