INDEX

Explanations

phrases related to a specific geographical location, "Pol"

oai_token-act-pair · gpt-3.5-turbo

references to locations and entities associated with the term "Pol."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

IELD

-0.82

URN

-0.77

ACTED

-0.76

TEXTURE

-0.74

ORGE

-0.73

ULTS

-0.73

OTAL

-0.73

é¾įåĸļå£«

-0.69

 bounds

-0.68

[|

-0.68

POSITIVE LOGITS

ipop

1.13

ikarp

1.05

itely

0.98

anco

0.95

atility

0.94

anski

0.88

itary

0.87

iov

0.84

ander

0.83

atile

0.83

Activations Density 0.007%

phrases related to a specific geographical location, "Pol"

references to locations and entities associated with the term "Pol."

No Comments

No Known Activations