INDEX

Explanations

phrases related to activities or events happening behind the scenes or in secret

oai_token-act-pair · gpt-3.5-turbo

references to behind-the-scenes activities

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

RIC

-0.75

arily

-0.69

 msec

-0.62

ERG

-0.61

ital

-0.60

lier

-0.58

pring

-0.58

 fined

-0.57

estate

-0.57

aez

-0.57

POSITIVE LOGITS

 scenes

1.37

 curtain

1.20

 Scenes

1.17

 curtains

0.96

scenes

0.91

 curve

0.90

 veil

0.81

 wheel

0.81

 lines

0.72

facade

0.71

Activations Density 0.097%

phrases related to activities or events happening behind the scenes or in secret

references to behind-the-scenes activities

No Comments

No Known Activations