INDEX

Explanations

words and phrases related to remaining in a specific position or state

oai_token-act-pair · gpt-3.5-turbo

references to continuity and persistence in situations or practices

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

apest

-0.86

 converter

-0.67

romy

-0.65

iscons

-0.64

ãĥ¯ãĥ³

-0.59

rocal

-0.59

VS

-0.58

eco

-0.57

imilar

-0.57

romancer

-0.57

POSITIVE LOGITS

 indefinitely

1.56

 until

1.53

 despite

1.43

 till

1.38

until

1.38

despite

1.31

 unchanged

1.21

 notwithstanding

1.15

withstanding

1.11

 regardless

1.11

Activations Density 0.738%

words and phrases related to remaining in a specific position or state

references to continuity and persistence in situations or practices

No Comments

No Known Activations