INDEX

Explanations

contractions of words with specific characters such as 'n't'

oai_token-act-pair · gpt-3.5-turbo

instances of the word "wouldn't" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

æ³

-0.68

ARM

-0.68

story

-0.66

ULT

-0.63

PI

-0.62

 Case

-0.62

PU

-0.61

ocal

-0.60

Adv

-0.60

agency

-0.60

POSITIVE LOGITS

't

1.09

ģĸ

0.82

 never

0.79

terness

0.78

atically

0.76

geon

0.76

¹

0.74

 surely

0.73

¨

0.73

±

0.73

Activations Density 0.009%

contractions of words with specific characters such as 'n't'

instances of the word "wouldn't" in various contexts

No Comments

No Known Activations