INDEX

Explanations

phrases containing names of organizations or companies

oai_token-act-pair · gpt-3.5-turbo

instances of the word "is" in various contexts

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

owan

-0.82

izable

-0.76

 glances

-0.75

 Stones

-0.71

 Ends

-0.70

ipeg

-0.70

 highs

-0.69

 IMAGES

-0.69

 Telesc

-0.68

rouse

-0.68

POSITIVE LOGITS

 headquartered

1.34

 sponsoring

1.20

 responsible

1.09

 suing

1.08

 obligated

1.07

 cooperating

1.06

 affiliated

1.04

 aware

1.02

 embroiled

1.01

 complicit

1.01

Activations Density 0.321%

phrases containing names of organizations or companies

instances of the word "is" in various contexts

No Comments

No Known Activations