INDEX

Explanations

Roman numerals followed by an 's' or alternative variations thereof

oai_token-act-pair · gpt-3.5-turbo

occurrences of the letter 'L' in various contexts

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

 Pieces

-0.69

 Seym

-0.66

 Stam

-0.65

 Antar

-0.63

 Jinn

-0.63

 shenan

-0.62

 Pony

-0.62

 Vaugh

-0.61

gib

-0.61

 Healer

-0.61

POSITIVE LOGITS

ï¸ı

1.33

âĸł

1.05

shall

1.04

could

1.00

should

0.99

Â£

0.99

Â§

0.96

 denotes

0.92

said

0.91

âĢł

0.90

Activations Density 0.316%

Roman numerals followed by an 's' or alternative variations thereof

occurrences of the letter 'L' in various contexts

No Comments

No Known Activations