INDEX

Explanations

mentions of the word "pro" within text

oai_token-act-pair · gpt-3.5-turbo

the word "ro" in various contexts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Coulter

-0.82

 EntityItem

-0.75

 Sons

-0.73

ãĥ¼ãĥĨãĤ£

-0.71

 eminent

-0.68

 Titanic

-0.66

CBI

-0.64

 Catalyst

-0.63

ividual

-0.63

 Beir

-0.62

POSITIVE LOGITS

spective

1.13

bably

1.08

cks

1.05

ffen

1.04

active

1.04

blems

1.03

ffe

1.01

spection

0.99

ven

0.98

gyn

0.98

Activations Density 0.024%

mentions of the word "pro" within text

the word "ro" in various contexts

No Comments

No Known Activations