INDEX

Explanations

words related to exits or departures

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate movement or transitions out of a place or situation

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

avorite

-0.70

tyr

-0.67

 arsen

-0.64

 laureate

-0.61

 etched

-0.60

hedral

-0.59

uzzle

-0.58

 Eternity

-0.57

 turnover

-0.57

 examiner

-0.57

POSITIVE LOGITS

stretched

1.19

fitted

1.15

casts

1.07

smart

1.03

lander

1.01

doing

0.99

flows

0.98

lier

0.98

doors

0.97

bur

0.97

Activations Density 0.069%

words related to exits or departures

phrases that indicate movement or transitions out of a place or situation

No Comments

No Known Activations