INDEX

Explanations

mentions of the word "Hind" with varying degrees of specificity

oai_token-act-pair · gpt-3.5-turbo

references to the term "Hind" and related variations

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

ãĥ¯

-0.80

-0.75

ãĤ£

-0.73

ailability

-0.71

ç¥ŀ

-0.68

ynthesis

-0.68

è¦ļéĨĴ

-0.67

ITED

-0.66

ä»£

-0.64

uador

-0.63

POSITIVE LOGITS

sight

1.19

ering

0.94

quarters

0.90

shire

0.88

rance

0.85

 Hind

0.84

nesses

0.83

ness

0.78

wich

0.77

rums

0.75

Activations Density 0.015%

mentions of the word "Hind" with varying degrees of specificity

references to the term "Hind" and related variations

No Comments

No Known Activations