INDEX

Explanations

words related to something being easy, simple, or logical

oai_token-act-pair · gpt-3.5-turbo

the term "straightforward" and related descriptors of simplicity

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

akin

-0.77

psey

-0.75

aughters

-0.73

reen

-0.71

wives

-0.70

ingle

-0.69

zona

-0.67

roid

-0.65

otos

-0.64

olina

-0.63

POSITIVE LOGITS

ly

1.18

 straightforward

0.97

lly

0.97

LY

0.92

ity

0.81

\\\\\\\\

0.81

é¾

0.80

itarian

0.75

NESS

0.75

tons

0.74

Activations Density 0.016%

words related to something being easy, simple, or logical

the term "straightforward" and related descriptors of simplicity

No Comments

No Known Activations