INDEX

Explanations

calls to verify the user's humanity

oai_token-act-pair · gpt-3.5-turbo

words and phrases related to verification and validation processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

NetMessage

-0.75

alm

-0.73

cent

-0.68

enburg

-0.68

illard

-0.68

tnc

-0.67

leton

-0.67

etic

-0.67

corn

-0.67

ideshow

-0.67

POSITIVE LOGITS

 authenticity

0.95

 verification

0.86

 verified

0.78

 verifying

0.78

worthiness

0.77

ificate

0.76

iably

0.76

 validity

0.75

 verify

0.73

ifies

0.71

Activations Density 0.063%

calls to verify the user's humanity

words and phrases related to verification and validation processes

No Comments

No Known Activations