INDEX

Explanations

instances of a person or thing returning to a specific place or activity

oai_token-act-pair · gpt-3.5-turbo

instances of returning or coming back

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

holes

-0.81

ascript

-0.73

merce

-0.72

Secondly

-0.68

anka

-0.68

Enlarge

-0.66

selves

-0.65

gey

-0.65

erity

-0.64

croft

-0.64

POSITIVE LOGITS

 haunt

1.09

 prominence

1.07

 action

0.83

 contention

0.83

 obscurity

0.82

 academia

0.80

 consciousness

0.76

 earth

0.76

 court

0.74

 shore

0.74

Activations Density 0.063%

instances of a person or thing returning to a specific place or activity

instances of returning or coming back

No Comments

No Known Activations