INDEX

Explanations

the word "Sham" with different levels of activation

oai_token-act-pair · gpt-3.5-turbo

mentions of "Sham" or related terms 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 @ 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

 mell

-0.72

 Extrem

-0.67

ATED

-0.62

ioxide

-0.62

IBLE

-0.61

 explos

-0.60

 newsletters

-0.60

lihood

-0.59

 pancakes

-0.59

 Gerry

-0.59

POSITIVE LOGITS

rock

1.20

blers

1.19

eless

1.19

aji

0.94

arak

0.89

urai

0.86

atari

0.85

ble

0.84

ted

0.84

ulhu

0.84

Activations Density 0.053%

the word "Sham" with different levels of activation

mentions of "Sham" or related terms in various contexts

No Comments

No Known Activations