INDEX

Explanations

phrases related to making authoritative statements or declarations

oai_token-act-pair · gpt-3.5-turbo

instances of the word "pronounce" and related variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Pool

-0.73

ien

-0.70

pool

-0.69

 Prism

-0.64

¶

-0.64

Settings

-0.64

 Camer

-0.64

Tank

-0.64

Storage

-0.63

 Tanks

-0.63

POSITIVE LOGITS

 pronounce

3.79

 pronounced

2.15

 pron

2.11

 pronunciation

1.84

unci

1.54

unciation

1.44

 syll

1.42

 pronoun

1.41

pron

1.32

 spelled

1.32

Activations Density 0.027%

phrases related to making authoritative statements or declarations

instances of the word "pronounce" and related variations

No Comments

No Known Activations