INDEX

Explanations

words related to connections or associations between different entities or variables

oai_token-act-pair · gpt-3.5-turbo

references to associations or connections between different subjects or conditions

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

 Penguins

-0.65

illation

-0.65

stall

-0.63

 Tiger

-0.62

ypes

-0.62

Sev

-0.58

esa

-0.58

 Germ

-0.57

FUL

-0.56

 sidelines

-0.56

POSITIVE LOGITS

linked

0.91

edin

0.75

paren

0.74

irect

0.73

 statically

0.72

lez

0.70

 geographically

0.68

chain

0.68

imentary

0.67

 closely

0.66

Activations Density 0.035%

words related to connections or associations between different entities or variables

references to associations or connections between different subjects or conditions

No Comments

No Known Activations