INDEX

Explanations

phrases related to defending or protecting

oai_token-act-pair · gpt-3.5-turbo

references to self-defense and entities related to news or commentary

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

raltar

-0.80

ppo

-0.77

 Seym

-0.75

cember

-0.71

hod

-0.70

anwhile

-0.70

leased

-0.68

oga

-0.68

heny

-0.67

enna

-0.66

POSITIVE LOGITS

liest

0.80

nces

0.80

 Staples

0.77

================================================================

0.72

atures

0.71

 Ãľ

0.71

cheon

0.70

IOR

0.70

 Schwarzenegger

0.70

incial

0.69

Activations Density 0.024%

phrases related to defending or protecting

references to self-defense and entities related to news or commentary

No Comments

No Known Activations