INDEX

Explanations

statements or messages containing verbs in the past tense

oai_token-act-pair · gpt-3.5-turbo

phrases related to official statements or messages

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

Ã±o

-0.72

apolis

-0.70

 Goodman

-0.66

pload

-0.66

engeance

-0.64

idable

-0.64

avement

-0.64

ilogy

-0.63

ascal

-0.63

icho

-0.63

POSITIVE LOGITS

 aloud

1.35

just

1.05

dress

0.98

mitt

0.90

ahead

0.83

mill

0.81

ILY

0.80

printed

0.79

 comprehension

0.77

ied

0.76

Activations Density 0.027%

statements or messages containing verbs in the past tense

phrases related to official statements or messages

No Comments

No Known Activations