INDEX

Explanations

phrases or questions related to visual comparisons or hypothetical scenarios

oai_token-act-pair · gpt-3.5-turbo

phrases that inquire about appearances or states of being

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Americ

-0.70

Force

-0.59

tsy

-0.58

cipl

-0.57

arters

-0.55

resent

-0.55

 flurry

-0.55

Sharp

-0.55

helps

-0.54

ãĥĥãĥī

-0.54

POSITIVE LOGITS

lihood

0.81

 WITHOUT

0.81

liest

0.78

 inside

0.75

 beforehand

0.74

 without

0.74

 nowadays

0.73

 unto

0.73

 compared

0.72

 outside

0.71

Activations Density 0.045%

phrases or questions related to visual comparisons or hypothetical scenarios

phrases that inquire about appearances or states of being

No Comments

No Known Activations