INDEX

Explanations

words related to improvement, safety, competition, transparency, accountability, productivity, and diversity

oai_token-act-pair · gpt-3.5-turbo

concepts related to improvement, safety, and accountability across various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 cameo

-0.65

eger

-0.56

 slit

-0.55

 faint

-0.54

 ellipt

-0.54

 lump

-0.54

 bastard

-0.54

Mat

-0.54

cise

-0.53

 farewell

-0.53

POSITIVE LOGITS

 levels

0.90

 among

0.87

 amongst

0.84

among

0.79

levels

0.79

fulness

0.77

 across

0.77

itism

0.76

liness

0.75

 efficiency

0.74

Activations Density 0.367%

words related to improvement, safety, competition, transparency, accountability, productivity, and diversity

concepts related to improvement, safety, and accountability across various contexts

No Comments

No Known Activations