INDEX

Explanations

words related to describing actions or situations

oai_token-act-pair · gpt-3.5-turbo

instances of the word "is" or variations of the verb "to be."

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

 Highlights

-0.69

pedia

-0.69

 Raise

-0.66

sonian

-0.65

 Hitch

-0.65

inav

-0.64

 Letters

-0.64

âĢİ

-0.63

ths

-0.62

 Reports

-0.62

POSITIVE LOGITS

nt

1.20

 deemed

0.92

 supposed

0.88

 already

0.85

 treated

0.85

 able

0.85

 undet

0.84

 unable

0.84

 taken

0.83

 indistinguishable

0.81

Activations Density 0.387%

words related to describing actions or situations

instances of the word "is" or variations of the verb "to be."

No Comments

No Known Activations