INDEX

Explanations

phrases related to controversies, politics, and investigations

oai_token-act-pair · gpt-3.5-turbo

words related to emotional intensity and severity

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

 Emblem

-0.71

urst

-0.68

 Incarn

-0.67

FS

-0.65

GTA

-0.64

ABE

-0.62

USL

-0.61

wagen

-0.60

raints

-0.60

FML

-0.59

POSITIVE LOGITS

 versa

1.09

 alike

1.06

 accordingly

0.97

 thereafter

0.93

 importantly

0.90

SPONSORED

0.89

 consequ

0.89

 consequently

0.86

depending

0.82

 thereto

0.81

Activations Density 0.203%

phrases related to controversies, politics, and investigations

words related to emotional intensity and severity

No Comments

No Known Activations