INDEX

Explanations

phrases related to addressing issues or problems

oai_token-act-pair · gpt-3.5-turbo

phrases that highlight direct quotes or statements

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Vaugh

-0.68

 conj

-0.68

 listed

-0.66

 Pavel

-0.61

nown

-0.61

jri

-0.61

 reprint

-0.60

 scheduled

-0.59

 favor

-0.59

 printer

-0.58

POSITIVE LOGITS

Cause

1.28

Interstitial

1.10

Absolutely

1.05

Very

1.04

cause

1.00

Extreme

0.97

taboola

0.97

Really

0.95

Tw

0.93

Better

0.92

Activations Density 0.029%

phrases related to addressing issues or problems

phrases that highlight direct quotes or statements

No Comments

No Known Activations