INDEX

Explanations

sentences with the word "are" followed by adjectives or adverbs

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the existence or presence of entities and groups

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

irmation

-0.68

 Supports

-0.65

 unfocusedRange

-0.63

orthern

-0.62

ulence

-0.62

itiveness

-0.61

 Horizon

-0.61

abil

-0.60

iton

-0.60

iture

-0.60

POSITIVE LOGITS

 increasingly

1.08

 scrambling

1.06

 accustomed

1.06

 understandably

1.02

 clam

1.02

 encouraged

1.02

 often

0.98

 outraged

0.98

 fascinated

0.97

 dying

0.95

Activations Density 0.205%

sentences with the word "are" followed by adjectives or adverbs

phrases indicating the existence or presence of entities and groups

No Comments

No Known Activations