INDEX

Explanations

negatively impacting terms, especially related to health or politics

oai_token-act-pair · gpt-3.5-turbo

phrases related to short-term and long-term concepts, particularly in relation to time

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Rica

-0.78

Cry

-0.70

 Lanka

-0.67

 GEAR

-0.66

 Sharma

-0.64

KN

-0.64

SOL

-0.64

 physic

-0.63

 Reflect

-0.61

 Reviews

-0.60

POSITIVE LOGITS

issue

1.04

sized

1.02

word

0.97

headed

0.97

named

0.96

division

0.96

thing

0.96

sounding

0.95

same

0.94

powered

0.93

Activations Density 0.259%

negatively impacting terms, especially related to health or politics

phrases related to short-term and long-term concepts, particularly in relation to time

No Comments

No Known Activations