INDEX

Explanations

phrases related to news agencies or press organizations

oai_token-act-pair · gpt-3.5-turbo

references to the word "Ag" 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 @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 Jenner

-0.74

ãĥĥãĤ¯

-0.67

 jumper

-0.65

 Sparkle

-0.64

 casing

-0.63

 Cors

-0.62

Â·Â·

-0.62

Library

-0.62

 disposal

-0.59

 macro

-0.59

POSITIVE LOGITS

reements

1.51

reement

1.45

encies

1.38

reed

1.30

atha

1.21

endas

1.17

nostic

1.14

ripp

1.10

rees

1.08

ree

1.06

Activations Density 0.018%

phrases related to news agencies or press organizations

references to the word "Ag" in various contexts

No Comments

No Known Activations