INDEX

Explanations

words ending in 'es'

oai_token-act-pair · gpt-3.5-turbo

occurrences of the suffix "es"

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

è»

-0.80

ItemTracker

-0.74

amental

-0.69

SHIP

-0.69

ONSORED

-0.67

£ı

-0.67

icio

-0.67

iage

-0.67

ancial

-0.66

atform

-0.65

POSITIVE LOGITS

terday

1.24

ktop

1.16

peed

1.07

andro

1.06

earch

1.03

pace

0.95

bians

0.94

pec

0.93

oteric

0.91

lem

0.90

Activations Density 0.034%

words ending in 'es'

occurrences of the suffix "es"

No Comments

No Known Activations