INDEX

Explanations

phrases related to achievement or accolades

oai_token-act-pair · gpt-3.5-turbo

concepts related to organizational structures and events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

avour

-0.74

 civilisation

-0.72

 metre

-0.66

Britain

-0.66

uminium

-0.65

ãĤ¨ãĥ«

-0.63

 ageing

-0.63

Firstly

-0.61

colour

-0.60

 neighbours

-0.60

POSITIVE LOGITS

 supplemental

0.60

.)

0.58

 similarly

0.56

).[

0.56

.}

0.55

.]

0.54

 additional

0.53

NYC

0.52

NYU

0.51

 Indiana

0.51

Activations Density 2.075%

phrases related to achievement or accolades

concepts related to organizational structures and events

No Comments

No Known Activations