INDEX

Explanations

phrases related to technology and computer software version releases

oai_token-act-pair · gpt-3.5-turbo

words related to specific brands or trademarks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 compartment

-0.61

 patents

-0.61

 allowances

-0.61

enegger

-0.60

stood

-0.60

 redundancy

-0.60

 stewards

-0.59

 fellows

-0.59

ById

-0.58

Ö¼

-0.58

POSITIVE LOGITS

liction

0.78

theless

0.77

achine

0.73

jong

0.73

atra

0.73

obe

0.70

 Settlement

0.70

thia

0.69

eus

0.69

ciating

0.68

Activations Density 0.067%

phrases related to technology and computer software version releases

words related to specific brands or trademarks

No Comments

No Known Activations