INDEX

Explanations

instances of the word "little" used in a figurative or qualitative sense

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate minor reservations or qualifiers

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ulkan

-0.86

idth

-0.85

psons

-0.77

abad

-0.76

aye

-0.74

iership

-0.73

assador

-0.71

 Flavoring

-0.71

atu

-0.69

querade

-0.69

POSITIVE LOGITS

 schizophren

1.04

odd

1.03

bit

1.03

 rusty

1.03

 scary

1.03

 tricky

1.02

 overboard

1.01

 strange

1.00

 shaky

1.00

 fuzzy

0.99

Activations Density 0.035%

instances of the word "little" used in a figurative or qualitative sense

phrases that indicate minor reservations or qualifiers

No Comments

No Known Activations