INDEX

Explanations

German words and phrases, possibly related to legal or bureaucratic contexts

oai_token-act-pair · gpt-3.5-turbo

terms related to statistical analysis or data representation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

sounding

-0.68

ially

-0.65

 Mayweather

-0.65

Previous

-0.64

 Coyotes

-0.61

 Livingston

-0.60

 Carey

-0.58

 Colombian

-0.58

 Fallon

-0.57

straight

-0.55

POSITIVE LOGITS

heit

1.42

von

1.20

der

1.08

mit

1.04

ung

1.04

hof

1.03

wagen

1.02

und

1.02

stadt

0.99

sow

0.97

Activations Density 0.105%

German words and phrases, possibly related to legal or bureaucratic contexts

terms related to statistical analysis or data representation

No Comments

No Known Activations