INDEX

Explanations

highly emotive or impactful words

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "words."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

roller

-0.75

ania

-0.68

cade

-0.68

iday

-0.66

farm

-0.64

involved

-0.64

TOR

-0.64

aper

-0.64

iary

-0.63

availability

-0.63

POSITIVE LOGITS

 words

3.85

 Words

2.71

words

2.32

Words

2.07

 word

2.04

 phrases

1.99

word

1.83

 phrase

1.76

 adject

1.70

 verbs

1.55

Activations Density 0.026%

highly emotive or impactful words

occurrences of the word "words."

No Comments

No Known Activations