INDEX

Explanations

phrases related to changes or adjustments being made multiple times

oai_token-act-pair · gpt-3.5-turbo

phrases indicating changes or transformations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Corpus

-0.61

inar

-0.59

 escort

-0.57

Timeout

-0.57

SSL

-0.55

winner

-0.55

loe

-0.55

colo

-0.55

ilege

-0.55

 Refugees

-0.54

POSITIVE LOGITS

 drastically

1.18

 considerably

1.09

 radically

1.04

 dramatically

1.02

 substantially

1.01

 significantly

0.94

 greatly

0.90

 markedly

0.85

 tremendously

0.83

 slightly

0.83

Activations Density 0.507%

phrases related to changes or adjustments being made multiple times

phrases indicating changes or transformations

No Comments

No Known Activations