INDEX

Explanations

words related to trying, attempting, and efforts

oai_token-act-pair · gpt-3.5-turbo

forms of the verb "try" and associated actions

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

 Serie

-0.68

 Uran

-0.67

lance

-0.65

imov

-0.65

ERO

-0.65

eers

-0.62

 Palest

-0.61

 Theodore

-0.61

-0.60

 Blumenthal

-0.60

POSITIVE LOGITS

culosis

0.90

 aback

0.81

anooga

0.74

obl

0.73

phrine

0.73

iterator

0.70

icult

0.67

unte

0.67

anie

0.66

cycles

0.66

Activations Density 0.022%

words related to trying, attempting, and efforts

forms of the verb "try" and associated actions

No Comments

No Known Activations