INDEX

Explanations

instances where something is described as clear or clarity is being emphasized

oai_token-act-pair · gpt-3.5-turbo

instances of the word "clear."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

eatures

-1.02

irl

-0.80

uld

-0.65

alez

-0.65

olt

-0.63

 rapp

-0.63

umat

-0.63

unte

-0.61

 Loft

-0.61

awks

-0.60

POSITIVE LOGITS

ances

1.16

 clear

1.07

clear

0.98

 clears

0.86

Clear

0.84

ance

0.83

 enough

0.80

 clearer

0.79

\\\\\\\\

0.76

Rect

0.76

Activations Density 0.024%

instances where something is described as clear or clarity is being emphasized

instances of the word "clear."

No Comments

No Known Activations