INDEX

Explanations

words related to specific styles or formats

oai_token-act-pair · gpt-3.5-turbo

references to various styles or categories of content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

atoon

-0.90

emi

-0.88

ipedia

-0.86

arus

-0.82

arte

-0.79

ora

-0.79

pedia

-0.78

wright

-0.77

ma

-0.76

omen

-0.75

POSITIVE LOGITS

quartered

0.77

 immersion

0.75

 dementia

0.74

 punishments

0.73

 confinement

0.72

 refinement

0.71

 landslide

0.68

lihood

0.67

 tendencies

0.67

 automatic

0.67

Activations Density 0.058%

words related to specific styles or formats

references to various styles or categories of content

No Comments

No Known Activations