INDEX

Explanations

instances where the concept of "elaborating" or providing detailed explanations is mentioned

oai_token-act-pair · gpt-3.5-turbo

instances of the word "elaborate" in various contexts

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

thood

-0.69

utral

-0.67

friend

-0.67

ocal

-0.66

mic

-0.64

hip

-0.64

aah

-0.63

ĭ

-0.63

zone

-0.63

cipled

-0.63

POSITIVE LOGITS

 elaborate

0.86

acebook

0.78

 explan

0.78

 corrid

0.74

 arrangements

0.73

 Explan

0.71

 arrang

0.71

 traged

0.69

 adolesc

0.67

sheets

0.67

Activations Density 0.024%

instances where the concept of "elaborating" or providing detailed explanations is mentioned

instances of the word "elaborate" in various contexts

No Comments

No Known Activations