INDEX

Explanations

instances of the word "writing" with varying emphasis in context

oai_token-act-pair · gpt-3.5-turbo

references to the act of writing

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

Ĭ±

-0.89

agara

-0.78

EGA

-0.78

illon

-0.74

azar

-0.73

abe

-0.72

rals

-0.70

rolet

-0.69

Afee

-0.67

 Shinra

-0.67

POSITIVE LOGITS

 poems

0.88

writing

0.88

smanship

0.82

 penned

0.82

writer

0.81

 essays

0.79

 poem

0.78

 poetry

0.76

 letters

0.76

 writing

0.75

Activations Density 0.040%

instances of the word "writing" with varying emphasis in context

references to the act of writing

No Comments

No Known Activations