INDEX

Explanations

phrases indicating comparison or contrast

oai_token-act-pair · gpt-3.5-turbo

references to contextual relationships and additional information in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

osponsors

-0.83

english

-0.76

ãĤ¨ãĥ«

-0.71

icons

-0.70

hops

-0.70

amia

-0.69

uay

-0.69

emale

-0.68

cycles

-0.68

conservancy

-0.66

POSITIVE LOGITS

0.93

,.

0.76

we

0.75

 statement

0.72

 scenario

0.72

 caveat

0.71

 affirmation

0.70

 realization

0.69

 logic

0.68

 limitation

0.68

Activations Density 0.093%

phrases indicating comparison or contrast

references to contextual relationships and additional information in the text

No Comments

No Known Activations