INDEX

Explanations

phrases or keywords related to expressing preferences

oai_token-act-pair · gpt-3.5-turbo

expressions of personal preference

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

breakers

-0.78

pack

-0.74

brance

-0.74

breaks

-0.74

effects

-0.73

gren

-0.72

angers

-0.68

breaker

-0.67

meta

-0.65

bleacher

-0.64

POSITIVE LOGITS

itism

0.75

rals

0.74

 Mistress

0.65

quickShipAvailable

0.65

yip

0.62

 ratios

0.61

ancy

0.61

 embodiment

0.61

 embodiments

0.60

 preferring

0.60

Activations Density 0.018%

phrases or keywords related to expressing preferences

expressions of personal preference

No Comments

No Known Activations