INDEX

Explanations

instances where a certain quantity is being compared or adjusted in relation to another quantity

oai_token-act-pair · gpt-3.5-turbo

phrase structures that compare options, particularly those using "instead of."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

Merit

-0.72

Values

-0.70

Reviewer

-0.67

é¾įåĸļå£«

-0.67

lli

-0.65

igent

-0.65

Enlarge

-0.63

WIND

-0.63

mare

-0.63

love

-0.63

POSITIVE LOGITS

 relying

1.19

 usual

0.98

 bothering

0.91

 letting

0.90

 staying

0.89

 having

0.86

 merely

0.85

 being

0.85

 necessarily

0.84

 traditional

0.84

Activations Density 0.113%

instances where a certain quantity is being compared or adjusted in relation to another quantity

phrase structures that compare options, particularly those using "instead of."

No Comments

No Known Activations