INDEX

Explanations

phrases indicating a strong or extreme contrast between different entities or situations

oai_token-act-pair · gpt-3.5-turbo

adjectives that indicate intensity or degree of contrast

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

irl

-0.69

uden

-0.69

yss

-0.66

yip

-0.65

 bandwagon

-0.63

 dearly

-0.62

ffen

-0.62

gem

-0.62

iren

-0.61

hirt

-0.60

POSITIVE LOGITS

 increments

1.23

 circumstances

1.06

 fashion

1.05

 quantities

1.01

 proportions

1.01

 terms

1.00

 accordance

0.95

 unison

0.93

 haste

0.92

 intervals

0.91

Activations Density 0.149%

phrases indicating a strong or extreme contrast between different entities or situations

adjectives that indicate intensity or degree of contrast

No Comments

No Known Activations