INDEX

Explanations

words related to proper names or titles

oai_token-act-pair · gpt-3.5-turbo

references to value or worth in various contexts

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

 unaff

-0.69

 traffickers

-0.68

vana

-0.64

 chronically

-0.60

meric

-0.60

 unequ

-0.60

GO

-0.60

 polar

-0.60

olar

-0.60

cyt

-0.60

POSITIVE LOGITS

shire

1.01

å§«

0.96

sburg

0.95

bridge

0.92

 Borough

0.92

 Abbey

0.91

tons

0.89

 Heath

0.87

hire

0.85

shed

0.84

Activations Density 0.117%

words related to proper names or titles

references to value or worth in various contexts

No Comments

No Known Activations