INDEX

Explanations

instances of inaccuracies or falsehoods in a text

oai_token-act-pair · gpt-3.5-turbo

terms related to deceptive or incorrect information and its consequences

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

negie

-0.77

oubted

-0.74

 GOODMAN

-0.71

ICA

-0.71

rolet

-0.70

verning

-0.70

ampions

-0.68

ificantly

-0.68

atorium

-0.67

interrupted

-0.66

POSITIVE LOGITS

 syndrome

1.01

glers

0.92

 Syndrome

0.85

 perpetrated

0.80

manship

0.73

 imaginable

0.69

Exception

0.68

ulence

0.67

mas

0.67

 practices

0.66

Activations Density 0.369%

instances of inaccuracies or falsehoods in a text

terms related to deceptive or incorrect information and its consequences

No Comments

No Known Activations