INDEX

Explanations

technical terms related to product features or characteristics

oai_token-act-pair · gpt-3.5-turbo

terms and phrases related to the usage and implications of technology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

erenn

-0.68

Category

-0.60

bledon

-0.59

ciating

-0.59

ividual

-0.58

pherd

-0.57

Eth

-0.57

hetti

-0.56

rison

-0.55

ograp

-0.53

POSITIVE LOGITS

ģĸ

0.62

)]

0.54

 Architects

0.53

 nutshell

0.53

 Malays

0.53

)?

0.53

 foreskin

0.52

 leaked

0.52

 Tradable

0.52

 comprises

0.51

Activations Density 0.323%

technical terms related to product features or characteristics

terms and phrases related to the usage and implications of technology

No Comments

No Known Activations