INDEX

Explanations

instances indicating decision-making or actions taken

oai_token-act-pair · gpt-3.5-turbo

instances of the word "after" indicating subsequent events or actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ãĥ³ãĤ¸

-0.91

tick

-0.86

bard

-0.82

aez

-0.76

oire

-0.75

 endif

-0.75

MENT

-0.75

orage

-0.74

âĨĳ

-0.74

Favorite

-0.74

POSITIVE LOGITS

 discovering

1.23

 receiving

1.09

 witnessing

1.02

 revelations

1.02

 concluding

1.02

 noticing

1.01

 encountering

1.00

 failing

0.98

 learning

0.98

 realizing

0.97

Activations Density 0.110%

instances indicating decision-making or actions taken

instances of the word "after" indicating subsequent events or actions

No Comments

No Known Activations