INDEX

Explanations

words related to online or digital interactions

oai_token-act-pair · gpt-3.5-turbo

references to interactions and dealings 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 @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

rophe

-0.73

atum

-0.72

acht

-0.70

 inch

-0.65

 Negro

-0.64

dir

-0.63

Beg

-0.63

 blaze

-0.63

Osc

-0.63

erenn

-0.62

POSITIVE LOGITS

 interactions

3.56

 encounters

2.02

 relationships

1.98

 interaction

1.85

 dealings

1.80

 conversations

1.79

 behaviors

1.65

 interacting

1.65

 reactions

1.64

 behaviours

1.60

Activations Density 0.009%

words related to online or digital interactions

references to interactions and dealings in various contexts

No Comments

No Known Activations