INDEX

Explanations

the word "host" in various contexts

oai_token-act-pair · gpt-3.5-turbo

terms related to television hosts

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

 Rite

-0.75

-0.74

arbon

-0.66

âĨĳ

-0.65

 braking

-0.65

CLASSIFIED

-0.64

YP

-0.62

 bilateral

-0.62

prints

-0.61

 progressing

-0.61

POSITIVE LOGITS

esses

1.20

ess

1.00

names

0.90

name

0.87

host

0.86

 host

0.83

ilities

0.82

 hosts

0.74

ility

0.72

lov

0.71

Activations Density 0.019%

the word "host" in various contexts

terms related to television hosts

No Comments

No Known Activations