INDEX

Explanations

mentions of the word "Tur" at varying intensities

oai_token-act-pair · gpt-3.5-turbo

the occurrences of the term "Tur" and its variations in different contexts

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

[&

-0.78

 Cruel

-0.69

ãģį

-0.68

ãģ®éŃĶ

-0.68

 largeDownload

-0.67

 pleas

-0.63

 Known

-0.63

 Safe

-0.62

 Incarnation

-0.62

 Wesley

-0.61

POSITIVE LOGITS

fing

1.11

bul

1.09

geon

1.09

moil

1.08

lette

1.06

kish

1.01

rible

1.00

anche

0.99

pees

0.99

iosity

0.98

Activations Density 0.009%

mentions of the word "Tur" at varying intensities

the occurrences of the term "Tur" and its variations in different contexts

No Comments

No Known Activations