INDEX

Explanations

proper nouns ending in '-que'

oai_token-act-pair · gpt-3.5-turbo

instances of the term "que" or related words emphasizing the importance of community or inclusion

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

inating

-0.81

uania

-0.81

inately

-0.77

iaries

-0.75

lessly

-0.75

sonian

-0.73

inators

-0.72

iary

-0.71

iband

-0.70

omething

-0.70

POSITIVE LOGITS

que

1.08

lde

0.82

erness

0.81

xit

0.79

edo

0.76

aque

0.76

ague

0.75

Ã±o

0.74

zon

0.74

odore

0.73

Activations Density 0.014%

proper nouns ending in '-que'

instances of the term "que" or related words emphasizing the importance of community or inclusion

No Comments

No Known Activations