INDEX

Explanations

phrases indicating a progression or advancement to a higher level or state

oai_token-act-pair · gpt-3.5-turbo

expressions related to advancing or elevating levels 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 @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ipation

-0.73

\-

-0.62

etta

-0.62

 leftover

-0.61

minster

-0.61

 credited

-0.61

oided

-0.60

 counted

-0.59

quartered

-0.59

 listed

-0.58

POSITIVE LOGITS

 heights

1.14

 extremes

1.09

 levels

0.97

 depths

0.90

 boiling

0.86

 level

0.86

 LEVEL

0.84

 realms

0.83

 proportions

0.81

 pinnacle

0.80

Activations Density 0.191%

phrases indicating a progression or advancement to a higher level or state

expressions related to advancing or elevating levels in various contexts

No Comments

No Known Activations