INDEX

Explanations

proper nouns indicating people's names or titles

oai_token-act-pair · gpt-3.5-turbo

instances of group dynamics and collaboration

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

itiz

-0.70

supp

-0.67

Supp

-0.66

uria

-0.62

rina

-0.60

Sat

-0.57

=/

-0.56

 obsc

-0.55

teness

-0.55

iferation

-0.55

POSITIVE LOGITS

 respectively

1.65

 jointly

1.16

together

1.07

 together

1.06

 alike

0.99

 respective

0.97

 yourselves

0.94

 Together

0.93

both

0.89

 apiece

0.88

Activations Density 0.495%

proper nouns indicating people's names or titles

instances of group dynamics and collaboration

No Comments

No Known Activations