INDEX

Explanations

locations or sides indicated by directional terms

oai_token-act-pair · gpt-3.5-turbo

references to spatial locations and directions

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

Bir

-0.80

Weather

-0.75

ixtures

-0.74

edIn

-0.71

Sov

-0.70

accompan

-0.68

ÃįÃį

-0.68

astical

-0.68

forestation

-0.67

sbm

-0.67

POSITIVE LOGITS

 corner

0.85

of

0.77

 thereof

0.77

 ridge

0.73

 circles

0.69

 quad

0.68

zip

0.66

 onwards

0.65

 flank

0.64

tein

0.64

Activations Density 0.138%

locations or sides indicated by directional terms

references to spatial locations and directions

No Comments

No Known Activations