INDEX

Explanations

phrases related to correctness, accuracy, or being precise

oai_token-act-pair · gpt-3.5-turbo

instances of the word "right" 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 @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

ĸļ

-0.81

arettes

-0.69

ipation

-0.66

riber

-0.65

ains

-0.64

Bei

-0.63

gat

-0.63

ulz

-0.62

arette

-0.62

irl

-0.62

POSITIVE LOGITS

eous

1.16

shore

0.80

 wing

0.77

 winger

0.75

ness

0.69

 smack

0.69

ocrin

0.68

headed

0.67

ened

0.67

haven

0.67

Activations Density 0.047%

phrases related to correctness, accuracy, or being precise

instances of the word "right" in various contexts

No Comments

No Known Activations