INDEX

Explanations

phrases related to causation or reasoning

oai_token-act-pair · gpt-3.5-turbo

repetitive instances of the word "the."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

onica

-0.77

isode

-0.72

rust

-0.70

Hon

-0.70

Joined

-0.70

CHAPTER

-0.70

xit

-0.69

tackle

-0.68

ovo

-0.68

aug

-0.66

POSITIVE LOGITS

 majority

1.30

 latter

1.18

 specifics

1.13

 particulars

1.07

 vast

1.06

 wording

1.06

 numbers

1.00

 feds

1.00

 slightest

0.97

 exact

0.96

Activations Density 0.334%

phrases related to causation or reasoning

repetitive instances of the word "the."

No Comments

No Known Activations