INDEX

Explanations

descriptive phrases indicating completion or progression

oai_token-act-pair · gpt-3.5-turbo

present tense forms of the verb "to be."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ornia

-0.68

osate

-0.64

 sidx

-0.61

 Dare

-0.61

 Explain

-0.60

onica

-0.60

iny

-0.59

 Lack

-0.59

ionics

-0.58

 Ples

-0.58

POSITIVE LOGITS

 usually

1.44

 typically

1.34

 often

1.34

 invariably

1.28

 seldom

1.25

 rarely

1.23

 sometimes

1.16

 always

1.15

often

1.13

 therefore

1.11

Activations Density 0.226%

descriptive phrases indicating completion or progression

present tense forms of the verb "to be."

No Comments

No Known Activations