INDEX

Explanations

terms related to physical locations or directions

oai_token-act-pair · gpt-3.5-turbo

references to locations or routes along specific pathways or borders

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

thood

-0.96

geant

-0.72

ãĥīãĥ©

-0.69

videos

-0.68

vation

-0.66

ateurs

-0.64

 psychologists

-0.64

avorite

-0.63

PLIC

-0.62

OUP

-0.61

POSITIVE LOGITS

 periphery

1.15

 sidelines

0.95

 path

0.93

 corridors

0.87

 route

0.87

 seams

0.86

 dotted

0.84

 axis

0.83

 circumference

0.81

 corridor

0.79

Activations Density 0.071%

terms related to physical locations or directions

references to locations or routes along specific pathways or borders

No Comments

No Known Activations