INDEX

Explanations

adjectives expressing strong emotions or opinions

oai_token-act-pair · gpt-3.5-turbo

expressions of disbelief or incredulity

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

ascript

-0.89

onga

-0.73

ãĤ¼ãĤ¦ãĤ¹

-0.73

restling

-0.72

mentation

-0.71

HAEL

-0.71

igion

-0.71

atson

-0.68

-0.67

inion

-0.66

POSITIVE LOGITS

bet

0.75

 bookmark

0.71

 innocuous

0.69

 contradiction

0.65

 favorably

0.63

paren

0.63

 Zeal

0.62

 quaint

0.62

 obligated

0.60

uously

0.60

Activations Density 0.130%

adjectives expressing strong emotions or opinions

expressions of disbelief or incredulity

No Comments

No Known Activations