INDEX

Explanations

phrases related to specific or unspecified situations or scenarios

oai_token-act-pair · gpt-3.5-turbo

references to various "things" that imply topics or concepts under discussion

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

avorite

-0.68

incinn

-0.63

 Telecommunications

-0.60

KM

-0.60

 Facility

-0.59

oufl

-0.59

asers

-0.57

ynski

-0.57

ãĥ©ãĥ³

-0.56

ped

-0.56

POSITIVE LOGITS

iverse

1.21

 happening

1.10

 happened

1.05

 happen

0.90

 happens

0.90

Else

0.83

 happ

0.82

ional

0.82

0.81

 else

0.81

Activations Density 0.044%

phrases related to specific or unspecified situations or scenarios

references to various "things" that imply topics or concepts under discussion

No Comments

No Known Activations