INDEX

Explanations

phrases related to software and technology functions or actions

oai_token-act-pair · gpt-3.5-turbo

phrases that include the word "in" indicating various contexts or settings

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

¿½

-0.69

govtrack

-0.68

wagon

-0.68

ãĤ©

-0.66

gyn

-0.65

hoe

-0.63

selling

-0.62

Reviewer

-0.60

esley

-0.60

ngth

-0.60

POSITIVE LOGITS

 seconds

1.15

 milliseconds

1.10

 minutes

1.09

 situ

0.94

 real

0.92

animate

0.92

 fractions

0.91

 batches

0.84

 lieu

0.82

 accordance

0.82

Activations Density 0.142%

phrases related to software and technology functions or actions

phrases that include the word "in" indicating various contexts or settings

No Comments

No Known Activations