INDEX

Explanations

phrases related to specific locations or settings

oai_token-act-pair · gpt-3.5-turbo

instances of the word "in" and its surrounding context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 channels

-0.74

aces

-0.67

 piles

-0.67

 lines

-0.65

 strains

-0.65

 departments

-0.64

 forecasts

-0.63

ategories

-0.63

 indexes

-0.62

 trillions

-0.61

POSITIVE LOGITS

 theirs

0.82

lier

0.77

animate

0.65

 behalf

0.64

 titled

0.63

efficiency

0.63

 which

0.62

 yours

0.62

vol

0.61

 Scarlet

0.61

Activations Density 0.649%

phrases related to specific locations or settings

instances of the word "in" and its surrounding context

No Comments

No Known Activations