INDEX

Explanations

phrases related to location and physical descriptions

oai_token-act-pair · gpt-3.5-turbo

references to positions or roles of individuals within a structured context, such as uniforms or rankings

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ply

-0.77

 lengths

-0.69

 guiActiveUnfocused

-0.68

anship

-0.68

issions

-0.67

ipher

-0.66

ocations

-0.62

enei

-0.61

cellaneous

-0.60

nces

-0.60

POSITIVE LOGITS

liest

0.81

ãĥĺ

0.71

 Dracula

0.69

otto

0.68

Tuc

0.68

 Seat

0.65

iest

0.64

Picture

0.64

 tallest

0.63

 Cecil

0.62

Activations Density 0.409%

phrases related to location and physical descriptions

references to positions or roles of individuals within a structured context, such as uniforms or rankings

No Comments

No Known Activations