INDEX

Explanations

words related to solving or addressing problems

oai_token-act-pair · gpt-3.5-turbo

terms related to goals, problems, and challenges

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

 Speedway

-0.60

ãĥ³ãĤ¸

-0.60

ãĤ¨ãĥ«

-0.57

pen

-0.57

 Faces

-0.57

brook

-0.56

joice

-0.56

 originals

-0.55

 veggies

-0.55

 Doors

-0.55

POSITIVE LOGITS

antage

0.75

cture

0.73

iatus

0.72

ality

0.71

icio

0.69

itism

0.69

isphere

0.67

iety

0.64

 firsthand

0.64

undrum

0.63

Activations Density 0.264%

words related to solving or addressing problems

terms related to goals, problems, and challenges

No Comments

No Known Activations