INDEX

Explanations

beverages, specifically highlighting tea and coffee

oai_token-act-pair · gpt-3.5-turbo

phrases that include references to beverages

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

abilities

-0.91

cms

-0.77

iets

-0.75

nw

-0.75

aucuses

-0.74

arily

-0.71

hra

-0.70

ahime

-0.68

 tremend

-0.68

çīĪ

-0.67

POSITIVE LOGITS

 glass

0.81

 cloth

0.81

 disinfect

0.78

 whisky

0.77

 freshly

0.75

 corn

0.74

ice

0.73

 Scotch

0.72

 plastic

0.72

 silver

0.72

Activations Density 0.143%

beverages, specifically highlighting tea and coffee

phrases that include references to beverages

No Comments

No Known Activations