INDEX

Explanations

words related to negative or potentially harmful events or actions

oai_token-act-pair · gpt-3.5-turbo

terms related to negative events or situations

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

 Shotgun

-0.65

ModLoader

-0.63

buquerque

-0.62

raltar

-0.61

 FANTASY

-0.59

 cannabinoid

-0.57

Ô

-0.57

racuse

-0.55

mbuds

-0.55

 bund

-0.55

POSITIVE LOGITS

.[

0.95

.'

0.87

.}

0.86

 itself

0.84

.</

0.84

'.

0.79

.<

0.79

.]

0.78

 beforehand

0.78

.''

0.75

Activations Density 0.599%

words related to negative or potentially harmful events or actions

terms related to negative events or situations

No Comments

No Known Activations