INDEX

Explanations

adjectives related to positive attributes or emotions

oai_token-act-pair · gpt-3.5-turbo

the word "super" 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 @ 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Seym

-0.70

edIn

-0.68

Reloaded

-0.67

llers

-0.66

 Anat

-0.65

 Franks

-0.64

 Muse

-0.63

 Granger

-0.63

Tud

-0.63

 Sending

-0.62

POSITIVE LOGITS

visor

1.15

imposed

1.10

nova

1.07

visory

0.96

powers

0.90

cedes

0.89

charged

0.89

charg

0.86

computer

0.86

iour

0.84

Activations Density 0.012%

adjectives related to positive attributes or emotions

the word "super" in various contexts

No Comments

No Known Activations