INDEX

Explanations

negative words or phrases related to criticism or judgment

oai_token-act-pair · gpt-3.5-turbo

negative and positive descriptors related to health, behavior, and attributes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

hower

-0.78

hander

-0.75

ioxide

-0.75

jen

-0.75

jam

-0.74

hemer

-0.72

ften

-0.71

aniel

-0.71

former

-0.70

hirt

-0.70

POSITIVE LOGITS

 intentions

1.24

 backgrounds

1.20

 ambitions

1.16

 aspirations

1.12

 ties

1.11

 qualities

1.11

 interests

1.11

 feelings

1.10

 histories

1.08

 motives

1.07

Activations Density 0.336%

negative words or phrases related to criticism or judgment

negative and positive descriptors related to health, behavior, and attributes

No Comments

No Known Activations