INDEX

Explanations

instances of individuals being banned or suspended from various activities or locations

oai_token-act-pair · gpt-3.5-turbo

phrases related to being banned or suspended from activities or places

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 imperfect

-0.77

strate

-0.73

 adversity

-0.73

perfect

-0.68

wr

-0.67

reset

-0.64

ipel

-0.64

 wisely

-0.64

 underdog

-0.64

 dashed

-0.64

POSITIVE LOGITS

 indefinitely

1.06

 flights

1.00

 entrances

0.93

 participation

0.93

 entry

0.90

 gatherings

0.89

 imports

0.87

 activities

0.86

 altogether

0.86

 tours

0.85

Activations Density 0.360%

instances of individuals being banned or suspended from various activities or locations

phrases related to being banned or suspended from activities or places

No Comments

No Known Activations