INDEX

Explanations

occurrences of the word "ud" with a high activation level

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of a specific name or term within the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

xon

-0.88

borough

-0.66

#$

-0.65

Constructed

-0.63

 breaths

-0.63

hazard

-0.62

Ń·

-0.62

 tides

-0.60

 suspense

-0.60

 regress

-0.59

POSITIVE LOGITS

geon

1.22

ud

1.12

geoning

1.10

imentary

1.09

ible

1.05

iamond

1.02

olph

1.02

elta

0.99

onymous

0.96

apest

0.95

Activations Density 0.009%

occurrences of the word "ud" with a high activation level

repeated mentions of a specific name or term within the text

No Comments

No Known Activations