INDEX

Explanations

phrases related to accomplishment and achievement

oai_token-act-pair · gpt-3.5-turbo

indications of audience satisfaction and reactions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

hovah

-0.73

habi

-0.68

mite

-0.66

ãĥ¯

-0.59

ctuary

-0.58

 awaits

-0.58

 Â®

-0.57

 Says

-0.55

otin

-0.54

 Adds

-0.54

POSITIVE LOGITS

ranged

0.68

 tended

0.67

 lacked

0.66

Was

0.61

 hadn

0.60

 outwe

0.59

 hoped

0.58

 smelled

0.57

 knew

0.57

 beforehand

0.56

Activations Density 2.573%

phrases related to accomplishment and achievement

indications of audience satisfaction and reactions

No Comments

No Known Activations