INDEX

Explanations

words related to big size or scale

oai_token-act-pair · gpt-3.5-turbo

references to significant quantities or scales

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

ices

-0.80

pai

-0.74

apple

-0.73

utenant

-0.73

onto

-0.71

Reviewed

-0.70

emy

-0.70

cling

-0.69

zen

-0.68

eve

-0.67

POSITIVE LOGITS

 amounts

1.17

 quantities

1.03

 chunks

0.96

 scale

0.96

 proportions

0.95

 sums

0.94

 disproportion

0.87

 amount

0.86

 swat

0.85

 undertaking

0.84

Activations Density 0.051%

words related to big size or scale

references to significant quantities or scales

No Comments

No Known Activations