INDEX

Explanations

people's names that sound like they might be from a manufacturing company

oai_token-act-pair · gpt-3.5-turbo

names of individuals associated with a specific industry or context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

cffff

-0.76

bably

-0.65

ailability

-0.64

etheless

-0.64

ccording

-0.63

olulu

-0.60

 teasp

-0.60

 eleph

-0.59

 proport

-0.58

 confir

-0.58

POSITIVE LOGITS

pez

0.64

 Heist

0.62

wright

0.61

ÃĹ

0.60

 himself

0.57

âĢĲ

0.57

ãĢĳ

0.56

ragon

0.56

vil

0.56

utils

0.55

Activations Density 0.127%

people's names that sound like they might be from a manufacturing company

names of individuals associated with a specific industry or context

No Comments

No Known Activations