INDEX

Explanations

phrases related to something being complete or total

oai_token-act-pair · gpt-3.5-turbo

phrases indicating completeness or totality related to various subjects

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

apter

-0.81

pton

-0.73

amus

-0.72

=-=-=-=-

-0.72

tch

-0.70

Interstitial

-0.70

akov

-0.67

 Beir

-0.66

apters

-0.66

ICAN

-0.66

POSITIVE LOGITS

erton

0.89

screen

0.83

ness

0.81

 throttle

0.81

 brunt

0.80

 complement

0.78

 extent

0.77

 fled

0.77

 disclosure

0.75

 responsibility

0.73

Activations Density 0.029%

phrases related to something being complete or total

phrases indicating completeness or totality related to various subjects

No Comments

No Known Activations