INDEX

Explanations

comparisons using "like"

oai_token-act-pair · gpt-3.5-turbo

comparisons or similarities expressed using the word "like."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

wright

-0.79

kins

-0.78

arte

-0.77

both

-0.77

ateful

-0.77

avering

-0.75

alle

-0.75

avers

-0.74

alf

-0.74

their

-0.73

POSITIVE LOGITS

lihood

1.87

 structure

1.00

 atmosphere

1.00

 behavior

0.92

 antics

0.89

 structures

0.89

 confinement

0.89

 arrangement

0.88

 consistency

0.86

 behaviour

0.86

Activations Density 0.085%

comparisons using "like"

comparisons or similarities expressed using the word "like."

No Comments

No Known Activations