INDEX

Explanations

phrases related to pressure and urgency

oai_token-act-pair · gpt-3.5-turbo

forms 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

IFrame

Link

Negative Logits

ourt

-0.71

edIn

-0.67

 Visit

-0.65

quin

-0.65

 Popular

-0.64

Introduced

-0.63

 Supports

-0.63

plin

-0.61

endor

-0.61

irk

-0.60

POSITIVE LOGITS

nt

0.88

 muted

0.75

 hereby

0.73

 definitely

0.72

not

0.68

 unmist

0.66

 unclear

0.66

 unsur

0.66

 shown

0.65

 indeed

0.65

Activations Density 0.394%

phrases related to pressure and urgency

forms of the verb "to be."

No Known Activations