INDEX

Explanations

instances where the concept of "larger" is mentioned in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to the concept of size or magnitude, specifically the term "larger"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Exile

-0.75

onso

-0.75

GP

-0.73

SR

-0.72

ILLE

-0.72

elf

-0.71

EP

-0.70

Blade

-0.70

eline

-0.70

DVD

-0.70

POSITIVE LOGITS

 sized

1.09

 than

1.08

than

0.97

 intestine

0.97

 eleph

0.96

 quantities

0.86

 Than

0.84

 diameter

0.82

 carniv

0.81

 hippocamp

0.81

Activations Density 0.008%

instances where the concept of "larger" is mentioned in various contexts

references to the concept of size or magnitude, specifically the term "larger"

No Comments

No Known Activations