INDEX

Explanations

locations or places described in the text

oai_token-act-pair · gpt-3.5-turbo

names related to specific locations and groups

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

phia

-0.85

=]

-0.85

*/(

-0.76

IGHT

-0.76

ãĥ¯ãĥ³

-0.76

etheless

-0.73

nesday

-0.72

Pwr

-0.70

 DISTR

-0.68

ãĥī

-0.67

POSITIVE LOGITS

alore

0.99

zhou

0.94

oing

0.92

jiang

0.92

aroo

0.89

lasses

0.88

 Xuan

0.87

omez

0.87

eman

0.87

yang

0.85

Activations Density 0.022%

locations or places described in the text

names related to specific locations and groups

No Comments

No Known Activations