INDEX

Explanations

verbs related to actions and manipulation

oai_token-act-pair · gpt-3.5-turbo

actions or processes related to manipulation or control of objects and concepts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Dracula

-0.71

Cah

-0.69

Phi

-0.67

 Horton

-0.66

rium

-0.66

 Suffolk

-0.66

Boh

-0.65

 Chin

-0.62

 Commissioners

-0.62

our

-0.59

POSITIVE LOGITS

 anytime

1.19

 easily

1.11

 anywhere

1.05

 cheaply

1.04

 safely

0.99

 anonymously

0.95

 individually

0.91

 remotely

0.89

 arbitrarily

0.88

 statically

0.87

Activations Density 0.227%

verbs related to actions and manipulation

actions or processes related to manipulation or control of objects and concepts

No Comments

No Known Activations