INDEX

Explanations

words related to actions or transformations

oai_token-act-pair · gpt-3.5-turbo

action verbs and concepts related to processes and transformations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 disappearing

-0.70

 belonging

-0.67

 Neutral

-0.64

 participating

-0.60

 fleeing

-0.58

 flowing

-0.58

 moving

-0.58

 remaining

-0.57

 exiting

-0.57

 Antar

-0.57

POSITIVE LOGITS

ifies

1.66

izes

1.64

itates

1.46

ulates

1.42

uates

1.35

iates

1.34

ends

1.32

tains

1.27

poses

1.26

ishes

1.23

Activations Density 0.351%

words related to actions or transformations

action verbs and concepts related to processes and transformations

No Comments

No Known Activations