INDEX

Explanations

words related to different versions or forms of something, especially when changing or improving

oai_token-act-pair · gpt-3.5-turbo

concepts related to iterations and incarnations

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

yer

-0.93

ello

-0.78

isse

-0.78

rib

-0.76

roph

-0.74

rop

-0.72

med

-0.70

akh

-0.70

rophic

-0.68

ened

-0.68

POSITIVE LOGITS

 iteration

1.12

 incarnation

1.11

 iterations

0.99

 thereof

0.88

 incarn

0.87

Versions

0.76

 Reborn

0.75

 Karin

0.74

ADRA

0.74

isSpecialOrderable

0.73

Activations Density 0.014%

words related to different versions or forms of something, especially when changing or improving

concepts related to iterations and incarnations

No Comments

No Known Activations