INDEX

Explanations

words related to risks, dangers, and warnings

oai_token-act-pair · gpt-3.5-turbo

references to risks associated with various contexts and situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ergy

-0.88

gdala

-0.78

Hat

-0.76

thumbnails

-0.71

owned

-0.71

MRI

-0.70

cle

-0.69

Assembly

-0.69

CLE

-0.67

 Kinnikuman

-0.67

POSITIVE LOGITS

 risks

1.08

afety

1.06

 hazards

0.97

 dangers

0.94

 pitfalls

0.91

 inherent

0.90

 endanger

0.89

 consequences

0.87

 unpredict

0.85

hift

0.84

Activations Density 0.022%

words related to risks, dangers, and warnings

references to risks associated with various contexts and situations

No Comments

No Known Activations