INDEX

Explanations

instances where individuals are participating in activities or events

oai_token-act-pair · gpt-3.5-turbo

instances of participation in events or activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

weak

-0.71

assets

-0.69

 sharp

-0.67

chi

-0.67

 slit

-0.67

 ceilings

-0.65

heres

-0.64

ħĭ

-0.64

eli

-0.63

 flared

-0.63

POSITIVE LOGITS

 guiActiveUn

0.90

 è£ıç

0.73

vik

0.70

_-_

0.69

sson

0.68

Oss

0.67

uary

0.67

OTA

0.67

advertisement

0.64

odox

0.64

Activations Density 0.015%

instances where individuals are participating in activities or events

instances of participation in events or activities

No Comments

No Known Activations