INDEX

Explanations

references to written or composed content like letters, blog posts, novels, articles, or songs

oai_token-act-pair · gpt-3.5-turbo

references to writing and types of written works

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ALE

-0.76

areth

-0.65

Ability

-0.65

ONSORED

-0.64

MSN

-0.64

 granted

-0.61

FACE

-0.61

 vantage

-0.61

seekers

-0.60

Location

-0.60

POSITIVE LOGITS

 screenplay

1.08

 memoir

1.03

 autobiography

0.96

itatively

0.94

 poem

0.91

 blog

0.90

 scathing

0.90

 diary

0.87

 poems

0.85

 letter

0.83

Activations Density 0.166%

references to written or composed content like letters, blog posts, novels, articles, or songs

references to writing and types of written works

No Comments

No Known Activations