INDEX

Explanations

names and locations related to an event or place named Khan

oai_token-act-pair · gpt-3.5-turbo

references to the location "Khan" in various contexts

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

ension

-0.73

iculty

-0.69

ources

-0.68

ection

-0.67

vable

-0.66

ISION

-0.65

ITION

-0.64

ezvous

-0.64

ensions

-0.63

FACE

-0.63

POSITIVE LOGITS

istani

1.02

 Khan

0.93

 Sheikh

0.88

patrick

0.86

regate

0.83

abal

0.83

wal

0.82

atin

0.80

kat

0.80

 Noon

0.79

Activations Density 0.009%

names and locations related to an event or place named Khan

references to the location "Khan" in various contexts

No Comments

No Known Activations