INDEX

Explanations

phrases indicating a clear understanding or revelation of a situation

oai_token-act-pair · gpt-3.5-turbo

instances of clarity or clear conclusions in statements

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

umbn

-0.70

eries

-0.66

izons

-0.66

avorite

-0.63

OVA

-0.61

eatures

-0.61

inqu

-0.61

pes

-0.61

sembly

-0.60

 passively

-0.59

POSITIVE LOGITS

 enough

0.81

why

0.77

aneously

0.74

cut

0.72

 that

0.70

:]

0.70

how

0.69

ances

0.67

 sailing

0.67

gat

0.66

Activations Density 0.034%

phrases indicating a clear understanding or revelation of a situation

instances of clarity or clear conclusions in statements

No Comments

No Known Activations