INDEX

Explanations

phrases indicating a final mention or conclusion

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "last but not least."

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

ijn

-0.67

iral

-0.67

perate

-0.67

arter

-0.66

asel

-0.65

borg

-0.65

ibel

-0.65

ocide

-0.65

inition

-0.64

uum

-0.62

POSITIVE LOGITS

Ù

0.73

 prominently

0.67

 notable

0.66

!--

0.66

 Ø§ÙĦ

0.65

 âĢº

0.64

 importantly

0.64

rand

0.64

 Cosmos

0.63

taining

0.63

Activations Density 0.005%

phrases indicating a final mention or conclusion

instances of the phrase "last but not least."

No Comments

No Known Activations