INDEX

Explanations

questions or statements followed by actions or intentions to be carried out

oai_token-act-pair · gpt-3.5-turbo

phrases that involve asking questions or addressing issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

aughed

-0.64

 condem

-0.62

been

-0.59

ãĤ¦ãĤ¹

-0.57

tips

-0.55

è¦

-0.53

\">

-0.53

Prev

-0.53

+.

-0.52

POSITIVE LOGITS

 requires

1.10

we

0.87

requires

0.83

 please

0.81

 oneself

0.77

 involves

0.77

you

0.76

 depends

0.75

 lies

0.69

 properly

0.69

Activations Density 0.244%

questions or statements followed by actions or intentions to be carried out

phrases that involve asking questions or addressing issues

No Comments

No Known Activations