INDEX

Explanations

words that convey acknowledgment or admission, particularly in the context of passwords and logins

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_1/width_16k/average_l0_102

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

quedo

-0.61

olino

-0.60

ecka

-0.59

;'>

-0.57

</thead>

-0.56

SanPham

-0.55

plication

-0.55

Viki

-0.54

 OFDb

-0.53

helf

-0.53

POSITIVE LOGITS

 rowspan

1.33

 admit

1.09

 Admit

0.82

 login

0.70

tagHelperRunner

0.69

Login

0.67

 Login

0.67

 argue

0.65

 admits

0.63

 argued

0.61

Activations Density 0.036%

words that convey acknowledgment or admission, particularly in the context of passwords and logins

No Comments

No Known Activations