INDEX

Explanations

phrases containing special characters indicating an important point or emphasis

oai_token-act-pair · gpt-3.5-turbo

occurrences of a specific symbol or marker

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 shroud

-0.76

 shack

-0.74

 touches

-0.71

 closest

-0.71

weights

-0.71

 dispers

-0.70

 nude

-0.69

 assemb

-0.68

 Downs

-0.68

 Borough

-0.68

POSITIVE LOGITS

¬

1.17

£

1.11

º

1.11

ľ

1.09

Ĵ

1.09

¹

1.07

«

1.07

ķ

1.06

¡

1.06

¥

1.05

Activations Density 0.139%

phrases containing special characters indicating an important point or emphasis

occurrences of a specific symbol or marker

No Comments

No Known Activations