INDEX

Explanations

keywords related to geographical or categorical areas

oai_token-act-pair · gpt-3.5-turbo

references to specific fields or contexts that indicate areas or matters requiring attention or consideration

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

iasis

-0.75

 Daytona

-0.65

issance

-0.64

 Bulldogs

-0.59

izons

-0.58

esi

-0.57

DE

-0.57

Es

-0.56

atorium

-0.55

ITAL

-0.55

POSITIVE LOGITS

afety

1.06

pread

0.97

mith

0.83

uggest

0.82

ometimes

0.81

 ranging

0.79

omething

0.79

pace

0.79

poon

0.79

pring

0.77

Activations Density 0.553%

keywords related to geographical or categorical areas

references to specific fields or contexts that indicate areas or matters requiring attention or consideration

No Comments

No Known Activations