INDEX

Explanations

adjectives describing a high level of something, such as "highly sought-after" or "highly respected."

oai_token-act-pair · gpt-3.5-turbo

the presence of the word "highly" and its variations

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

 Sins

-0.80

 Chaser

-0.79

 Hebdo

-0.77

 Slug

-0.77

 Tycoon

-0.76

 Ashe

-0.76

mere

-0.75

ories

-0.75

 Dare

-0.75

 Carry

-0.74

POSITIVE LOGITS

 unlikely

0.96

 regarded

0.93

 rated

0.90

 educated

0.90

 susceptible

0.89

 redundant

0.89

 insulated

0.88

 interconnected

0.88

 correlated

0.88

 influential

0.87

Activations Density 0.014%

adjectives describing a high level of something, such as "highly sought-after" or "highly respected."

the presence of the word "highly" and its variations

No Comments

No Known Activations