INDEX

Explanations

phrases related to looking or observing different subjects

oai_token-act-pair · gpt-3.5-turbo

references to societal observations or critiques

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

cil

-0.81

NULL

-0.71

)=(

-0.70

thens

-0.68

LEASE

-0.68

rave

-0.68

ahu

-0.67

rio

-0.66

erred

-0.66

Allows

-0.65

POSITIVE LOGITS

 totality

1.07

 histories

1.05

 archives

1.03

 wreckage

1.03

 literature

1.01

 history

0.99

 numbers

0.99

 demographics

0.99

 positives

0.99

 backgrounds

0.98

Activations Density 0.262%

phrases related to looking or observing different subjects

references to societal observations or critiques

No Comments

No Known Activations