INDEX

Explanations

mentions of relationships or connections between concepts or entities

oai_token-act-pair · gpt-3.5-turbo

plural forms of nouns

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

 elector

-0.66

 Britons

-0.60

 crates

-0.56

eur

-0.55

 euth

-0.54

 Overse

-0.54

 cipher

-0.53

 harshly

-0.52

 isolation

-0.52

 neoc

-0.52

POSITIVE LOGITS

ashtra

0.88

inki

0.83

ometime

0.80

omething

0.80

omew

0.80

aurus

0.78

forth

0.77

istance

0.73

ushi

0.73

abi

0.72

Activations Density 0.025%

mentions of relationships or connections between concepts or entities

plural forms of nouns

No Comments

No Known Activations