INDEX

Explanations

positive adjectives describing something favorable or beneficial

oai_token-act-pair · gpt-3.5-turbo

phrases that express positive evaluations or affirmations regarding various subjects

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

racuse

-0.79

ategory

-0.79

opers

-0.73

ipient

-0.73

hyde

-0.71

ancies

-0.70

eters

-0.69

olson

-0.67

ppe

-0.66

ardy

-0.66

POSITIVE LOGITS

 enough

1.05

enough

0.88

bye

0.81

 fodder

0.80

 consolation

0.78

 synergy

0.78

 news

0.77

 surpr

0.77

 luck

0.76

 Enough

0.76

Activations Density 0.114%

positive adjectives describing something favorable or beneficial

phrases that express positive evaluations or affirmations regarding various subjects

No Comments

No Known Activations