INDEX

Explanations

phrases or words that are identified as particularly important or significant

oai_token-act-pair · gpt-3.5-turbo

instances of the word "particularly" and its variations used to emphasize importance or significance

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

ylon

-0.75

ruary

-0.67

glass

-0.66

CT

-0.66

glers

-0.65

Runner

-0.65

istor

-0.65

 demolition

-0.64

ORTS

-0.64

article

-0.64

POSITIVE LOGITS

 susceptible

0.99

 suited

0.98

 egregious

0.97

 noticeable

0.89

 vulnerable

0.89

 noteworthy

0.88

 advantageous

0.88

 acute

0.85

 concentrated

0.81

 situated

0.80

Activations Density 0.034%

phrases or words that are identified as particularly important or significant

instances of the word "particularly" and its variations used to emphasize importance or significance

No Comments

No Known Activations