INDEX

Explanations

actions related to covering something or someone

oai_token-act-pair · gpt-3.5-turbo

instances of the word "cover" and its variations in various contexts

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 Sabha

-0.80

 pleasing

-0.71

LV

-0.68

erers

-0.66

STEM

-0.65

shi

-0.65

arb

-0.65

GBT

-0.64

âĹ¼

-0.64

----------

-0.63

POSITIVE LOGITS

alls

1.04

ages

0.99

 cover

0.86

 umbrella

0.78

ity

0.77

 blanket

0.75

cover

0.71

 covering

0.70

AGE

0.69

 gaps

0.69

Activations Density 0.030%

actions related to covering something or someone

instances of the word "cover" and its variations in various contexts

No Known Activations