INDEX

Explanations

verbs related to recommendations or predictions

oai_token-act-pair · gpt-3.5-turbo

the word "be" used in various contexts, indicating states or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Inquis

-0.70

 Awakens

-0.69

atures

-0.66

terday

-0.64

 Franch

-0.64

 corners

-0.64

orers

-0.61

aware

-0.60

 Norn

-0.60

 Moose

-0.59

POSITIVE LOGITS

 considered

1.07

 regarded

1.06

 judged

1.04

 construed

1.04

 viewed

1.02

 interpreted

1.02

 deemed

0.97

 debated

0.96

 reviewed

0.96

 handled

0.95

Activations Density 0.253%

verbs related to recommendations or predictions

the word "be" used in various contexts, indicating states or conditions

No Comments

No Known Activations