INDEX

Explanations

contexts or situations that are crucial for understanding people or things

oai_token-act-pair · gpt-3.5-turbo

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

STON

-0.89

deen

-0.77

riage

-0.76

DAQ

-0.74

vals

-0.73

owship

-0.73

thing

-0.67

CVE

-0.66

doms

-0.65

ãĥīãĥ©

-0.64

POSITIVE LOGITS

ually

1.04

ual

1.00

uality

0.91

 context

0.90

 Context

0.76

 contexts

0.76

 contextual

0.74

 menus

0.72

 menu

0.70

ful

0.70

Activations Density 0.031%

contexts or situations that are crucial for understanding people or things

No Comments

No Known Activations