INDEX

Explanations

phrases related to negative events happening to physical locations like homes and buildings

oai_token-act-pair · gpt-3.5-turbo

references to homes or buildings involved in incidents or damage

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

 congratulations

-0.78

 strokes

-0.75

verse

-0.75

 phrases

-0.73

 comparisons

-0.73

 phrase

-0.70

 wavelengths

-0.68

 motions

-0.68

zi

-0.67

 waves

-0.65

POSITIVE LOGITS

 reopened

1.15

 housed

1.04

 renovated

1.00

 situated

1.00

 remod

0.99

 demolished

0.99

 flooded

0.97

 emptied

0.96

Built

0.96

 erected

0.93

Activations Density 0.449%

phrases related to negative events happening to physical locations like homes and buildings

references to homes or buildings involved in incidents or damage

No Comments

No Known Activations