INDEX

Explanations

words related to potential negative outcomes or dangers

oai_token-act-pair · gpt-3.5-turbo

references to risks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

ople

-0.71

iture

-0.62

 Favorite

-0.62

imon

-0.62

ition

-0.62

otion

-0.61

cel

-0.61

 Friend

-0.60

orate

-0.59

ixt

-0.58

POSITIVE LOGITS

 risks

3.84

 risk

2.57

 dangers

2.17

 hazards

2.09

risk

2.02

 Risk

1.92

 risked

1.73

 harms

1.73

 disadvantages

1.61

 pitfalls

1.60

Activations Density 0.020%

words related to potential negative outcomes or dangers

references to risks

No Comments

No Known Activations