INDEX

Explanations

phrases indicating causality or contribution

oai_token-act-pair · gpt-3.5-turbo

phrases that discuss causality and relationships between elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

apse

-0.76

aby

-0.70

lege

-0.69

VI

-0.69

scribe

-0.64

must

-0.64

anything

-0.63

æ³

-0.63

aii

-0.62

apolis

-0.60

POSITIVE LOGITS

 demographics

0.78

 anecd

0.74

 Flavoring

0.71

 sheer

0.68

 misunder

0.68

 reluctance

0.67

 inertia

0.66

 awareness

0.64

 implicit

0.64

 underest

0.63

Activations Density 0.144%

phrases indicating causality or contribution

phrases that discuss causality and relationships between elements

No Comments

No Known Activations