INDEX

Explanations

phrases indicating high importance or necessity

oai_token-act-pair · gpt-3.5-turbo

phrases emphasizing necessity or strong recommendations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

vironment

-0.87

vertisements

-0.75

vernment

-0.72

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.71

ividual

-0.71

hyde

-0.68

foreseen

-0.67

iency

-0.64

gdala

-0.64

 Hamp

-0.64

POSITIVE LOGITS

hew

0.79

ANG

0.78

aches

0.78

angs

0.77

 reproduce

0.77

 abide

0.76

reads

0.75

afa

0.74

ards

0.72

sis

0.70

Activations Density 0.024%

phrases indicating high importance or necessity

phrases emphasizing necessity or strong recommendations

No Comments

No Known Activations