INDEX

Explanations

words related to physical attributes or characteristics

oai_token-act-pair · gpt-3.5-turbo

negative phrases or contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 Expend

-0.72

 materially

-0.64

 opin

-0.64

 Patreon

-0.64

 Emmy

-0.63

RBI

-0.63

 intel

-0.62

 Instr

-0.62

 Fiscal

-0.61

itÃ©

-0.61

POSITIVE LOGITS

based

1.35

shaped

1.29

like

1.15

type

1.13

sized

1.13

length

1.12

style

1.11

mounted

1.08

plate

1.08

clad

1.07

Activations Density 0.102%

words related to physical attributes or characteristics

negative phrases or contexts

No Comments

No Known Activations