INDEX

Explanations

phrases used in direct speech to continue a point

oai_token-act-pair · gpt-3.5-turbo

instances of the word "continued."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

arta

-0.93

errilla

-0.76

yip

-0.73

ilibrium

-0.70

eros

-0.69

ousands

-0.67

ocene

-0.65

assium

-0.60

odor

-0.60

ids

-0.58

POSITIVE LOGITS

0.95

:"

0.93

 recount

0.88

 stating

0.87

 omin

0.86

 explaining

0.86

 asserting

0.84

 saying

0.83

0.78

:]

0.78

Activations Density 0.045%

phrases used in direct speech to continue a point

instances of the word "continued."

No Comments

No Known Activations