INDEX

Explanations

phrases related to directives or imperatives

oai_token-act-pair · gpt-3.5-turbo

phrases that express safety and caution

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Interstitial

-0.82

Eva

-0.77

$.

-0.72

SourceFile

-0.69

Latest

-0.67

âĸĪ

-0.67

Untitled

-0.65

CV

-0.65

 Instr

-0.63

 ItemLevel

-0.62

POSITIVE LOGITS

':

0.84

?:

0.84

 meanwhile

0.79

'?

0.73

 looms

0.72

 aside

0.71

 Edit

0.68

huh

0.68

!:

0.67

 campaigners

0.67

Activations Density 0.935%

phrases related to directives or imperatives

phrases that express safety and caution

No Comments

No Known Activations