INDEX

Explanations

definitions or descriptions within a text

oai_token-act-pair · gpt-3.5-turbo

instances of the term "definitions" and related terminology

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

ajo

-0.72

hop

-0.71

hops

-0.67

pal

-0.64

sold

-0.63

wash

-0.63

email

-0.62

 Seat

-0.62

 nearby

-0.62

oyal

-0.62

POSITIVE LOGITS

 definitions

3.70

 definition

2.65

 Definitions

2.33

 Definition

2.10

Definition

1.82

definition

1.77

 meanings

1.75

 descriptions

1.56

initions

1.55

 defines

1.53

Activations Density 0.020%

definitions or descriptions within a text

instances of the term "definitions" and related terminology

No Comments

No Known Activations