INDEX

Explanations

words related to physical structures or objects, such as "cathedral," "circuits," "banner," and "robot."

oai_token-act-pair · gpt-3.5-turbo

nouns related to geographical features and structures

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

actionDate

-0.60

 Saying

-0.59

 Roses

-0.57

Vote

-0.57

inion

-0.57

Lear

-0.56

perate

-0.56

 Loss

-0.55

 declining

-0.54

relevant

-0.54

POSITIVE LOGITS

 boasts

1.15

 resembles

1.12

 combines

1.09

 represents

1.08

 comprises

1.06

 possesses

1.05

 debuted

1.05

 embodies

1.05

 incorporates

1.00

 delivers

0.98

Activations Density 0.393%

words related to physical structures or objects, such as "cathedral," "circuits," "banner," and "robot."

nouns related to geographical features and structures

No Comments

No Known Activations