INDEX

Explanations

numerical step instructions

oai_token-act-pair · gpt-3.5-turbo

step sequences or instructional elements

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

soDeliveryDate

-0.74

 nineteen

-0.69

atan

-0.65

atorial

-0.60

 warr

-0.60

athered

-0.60

 apprehension

-0.59

 unpredictable

-0.58

 hungry

-0.57

 interfere

-0.57

POSITIVE LOGITS

1.15

:{

0.99

0.90

:]

0.87

 Recap

0.86

:(

0.84

 Edit

0.82

.:

0.82

:-

0.82

 summarizes

0.80

Activations Density 0.107%

numerical step instructions

step sequences or instructional elements

No Comments

No Known Activations