INDEX

Explanations

phrases related to making something stronger or more robust

oai_token-act-pair · gpt-3.5-turbo

concepts related to enhancement or improvement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Marathon

-0.69

Hop

-0.66

 Gorge

-0.66

Mill

-0.66

ilk

-0.64

 Hallow

-0.62

 Nurse

-0.61

erenn

-0.61

rea

-0.60

POSITIVE LOGITS

ament

0.88

ments

0.83

 tremend

0.80

nesses

0.79

 strengthening

0.78

 bonds

0.78

ively

0.77

ifying

0.76

 streng

0.75

 horizont

0.74

Activations Density 0.016%

phrases related to making something stronger or more robust

concepts related to enhancement or improvement

No Comments

No Known Activations