INDEX

Explanations

phrases related to a sequence of actions or processes

oai_token-act-pair · gpt-3.5-turbo

phrases related to trends and significant findings

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

tics

-0.70

ancers

-0.69

venants

-0.68

 mosqu

-0.65

ournals

-0.63

 portfolios

-0.63

inventoryQuantity

-0.62

ournal

-0.62

azines

-0.62

 magazines

-0.62

POSITIVE LOGITS

 unthinkable

0.81

 echoed

0.81

 repeated

0.80

 sorely

0.77

ãĥĩãĤ£

0.75

albeit

0.74

 unimaginable

0.74

ldom

0.72

 envy

0.72

 seldom

0.71

Activations Density 0.380%

phrases related to a sequence of actions or processes

phrases related to trends and significant findings

No Comments

No Known Activations