INDEX

Explanations

phrases related to organizations and help

oai_token-act-pair · gpt-3.5-turbo

notable television shows or characters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 dispos

-0.68

aughtered

-0.65

 committing

-0.63

 speaking

-0.62

 whiff

-0.61

 Columb

-0.60

osing

-0.60

 Quentin

-0.59

 Virgin

-0.59

inh

-0.59

POSITIVE LOGITS

ãĥ¼ãĤ¯

0.71

 Tycoon

0.69

rative

0.69

ı

0.68

ãĤ¤ãĥĪ

0.67

democratic

0.67

slaught

0.66

moderate

0.66

Îº

0.65

temp

0.65

Activations Density 0.000%

phrases related to organizations and help

notable television shows or characters

No Comments

No Known Activations