INDEX

Explanations

quotes or attributed statements enclosed in quotation marks

oai_token-act-pair · gpt-3.5-turbo

phrases indicating sources or citations in text

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

 laure

-0.66

à¨

-0.65

abiding

-0.64

channelAvailability

-0.63

 discard

-0.63

acea

-0.62

":"/

-0.60

circle

-0.60

 coffers

-0.60

otin

-0.59

POSITIVE LOGITS

imes

0.70

iott

0.70

itcher

0.67

 Radio

0.67

 Observer

0.64

illi

0.64

iasm

0.63

icago

0.62

 Correspond

0.60

inen

0.60

Activations Density 0.218%

quotes or attributed statements enclosed in quotation marks

phrases indicating sources or citations in text

No Comments

No Known Activations